A LINGUISTIC ATLAS OF EARLY MIDDLE ENGLISH




INTRODUCTION


PART II: THE CORPUS


CHAPTER 4

TAGGING


Margaret Laing and Roger Lass




4. Tagging


We can best defend the scientific vitality of taxonomy by asserting ... that all systems of classification must express theories about the causes of order, and must therefore feature a complex mixture of concepts and percepts -- that is, preferences in human thinking combined with observations of nature’s often cryptic realities. Good taxonomies may be analogized with useful maps, but they reveal (as do all good maps) both our preferred mental schemes and the pieces of external reality that we have chosen to order and depict in our cartographic effort.

Gould, 2002: 289­–290.

4.1 The theory of tags
This atlas is based on a corpus of ‘lexico-grammatically tagged’ texts. It is necessary to define these terms: ‘lexico-grammatically’ may be unfamiliar, and ‘tagged’ has a number of possible interpretations. There are many styles of tagging, with different theoretical bases. Tagging is a taxonomic procedure, defining order and structure within the domain of texts. The tagged texts in LAEME primarily constitute a database for an atlas; they are also a quarry for other forms of linguistic study, e.g. orthographical, phonological, morphological and syntactic. But their primary purpose is to establish a taxonomy for grouping ‘items’ (sensu lato words and morphemes) in order to be able to compare and contrast them spatially and temporally.
A tag is a set of coordinates in a multidimensional space. Tags serve as addresses in this space, enabling us to locate analytically tractable objects, so that they can be extracted for processing. The two primary coordinates of tag-space are lexico-semantic identity and grammatical function. Both are themselves complex subspaces. Lexico-semantic and grammatical spaces are, or can be, hierarchical: certain spaces, given the nature of linguistic structure, are bound by definition to be subspaces of others. Thus functional space (containing categories like ‘direct object’) and subcategory space (e.g. ‘past tense’, ‘plural’) are dimensions of part-of-speech space, while subsenses of lexical items (e.g. ‘locative’, ‘temporal’) are dimensions of lexical or semantic space.
As we conceive tag-space, every item has grammatical coordinates, but not every item has lexical coordinates: e.g. pronouns, determiners and inflectional affixes are not given lexical labels, because they can be construed as carrying only grammatical information. In this we follow the common intuition that ‘grammatical’ forms are bound and lexical forms are free; the exceptions are pronouns and determiners, which though in languages like English are free forms, nonetheless carry only prototypically grammatical information, i.e. they code categories like person, number, gender, definiteness, deixis (for discussion see Croft 2003: §7.2.4). Notationally, the most extended tag type consists of a lexical element (‘lexel’) and a grammatical element (‘grammel’). Some tags, as indicated above, may consist of a grammel only, but none of a lexel only. An exemplary lexel/grammel pair would be:

$brother/nOd

where $ marks the beginning of a tag for computational purposes, and / separates the lexel and grammel, n = ‘noun’, O = ‘object’, d = ‘direct’. The lexel contains identifying semantic information (or to put it another way serves as a mnemonic: see below), and the grammel contains part-of-speech information (n ‘noun’), and functional information (n is an internal argument of its clause, serving the function ‘direct object’). A grammel may also indicate scope or operator/operand relations, e.g.

$brother/n<pr

where the argument is in the scope of the operator: pr = ‘preposition’. We define our tagging as ‘lexicogrammatical’, but in cases like this there is some syntactic tagging as well: the formulation above says that $brother is inside a prepositional phrase, in particular that it is governed by a preposition (i.e. is not an adjunct). This indication of ultra-word constituents (here ‘prepositional phrase’) is in a sense epiphenomenal, just as the label cj ‘conjunction’ is intended to mark a word but also indicates a clause or phrase boundary. Our tagging does not intend, in the first instance, to mark such higher-order constituents (though subsequent user-defined tagging at different levels may do so).1 The reason for marking operator/operand relations in cases like prepositional phrases is to facilitate recovery of information that may be of historical and/or regional significance. For instance, prepositional objects in early Middle English may be case-coded, and this is of historical and comparative interest. Similarly, at least in early texts, subordinate clauses may have different word-order properties from main clauses, and the grammel ‘cj’ serves as a flag for isolating such clauses for analysis.

A typical example of a grammel-only tag would be

$/P12N

which marks forms of ‘thou’ in subject position, where P = ‘personal pronoun’, 1 = ‘singular’, 2 = ‘second person’, N = ‘nominative’.


4.2 The status of tag elements
A tag is a theoretically heterogeneous object. The nonce-terms lexel and grammel have been invented deliberately, to avoid the potential theoretical baggage (as well as definitional specificity) carried by terms like ‘lexeme’, ‘lexical morpheme’ and ‘grammatical morpheme’. Our terminology is designed not to belong to any particular theoretical tradition, but to be compatible with whatever tradition(s) any user happens to work in or be familiar with. Lexels and grammels have different theoretical status. A lexel is in principle an atheoretical taxonomic convenience, merely a name given to an item. It is a language-specific mnemonic or identifier, and the choice of lexel does not attempt to characterise the structure of the lexicon. A lexel therefore makes no ontological assertions.2 In contrast, a grammel is ‘real’, in that a label like ‘n’ or ‘pr’ represents an object ‘presumed to exist’ within the structure of a language, and to be independently recognisable by outside observers. It makes a potentially realist claim about grammatical structure, within the ambit of a universalist (if limited) theory of grammar. We assume that ‘preposition’ or ‘noun’ or ‘conjunction’ are labels for objects that have non-arbitrary definitions within the universe of grammatical description. The labels are not text-specific: they are not restricted to any regional or temporal provenance. There is also, however, an element of projectible realism for at least some lexels. One would like to be able to ask questions like ‘what is the word for concept X in some other corpus?’. For instance, while $brother is intended merely as an identifier for a set of forms in our corpus, we would assume (a) that it would be related to another lexel — $sister, and (b) that any corpus we worked with would potentially have concepts that might be similarly identified.
Here is a characteristic listing of the forms (in upper case) associated with a tag, though in this example only the lexel is at issue:

$self/aj SELF, SELUE, SELUEN, SELUIN, SEOLF, SEOLFE, SEOLUE, SILF, SOLF, SUELF, SUF, SULF, SULFE

Such a display would suggest at first that the concept ‘lexel’ is really equivalent to ‘lexeme’; but this is not the case, even if in many instances they overlap. Consider, for instance, the main form-types3 associated with another tag:

$until/cj AL-FORT, AL-HwAT, AyAT, BITUIX-AND, BITUIX-AND-TIL, FORT, FOR-TO, FORTO-yAT, FOR-y~, FORd-y~, SO-yAT, SO-LONGE-yAT, TIL, TIL-yAT, TO, yAT

We would certainly not want to call these ‘word forms of the same lexeme’ (or for that matter ‘allomorphs of the same morpheme’). Historically and synchronically they simply represent different forms that were usable for the same semantic content. This is what we mean by characterising a lexel as a mnemonic rather than as a linguistic element. Lexels therefore may be of a number of different types, some dependent on historical contingencies, others on the linguistic level at which the tagger chooses to work. The tagger’s choice may be dependent on possible regional differences in the expression of the same semantic content. This for example explains the heterogeneity of forms under the semantic lexel $until above.


4.3. Typology of lexels
LAEME lexels are drawn from six different linguistic sources:

1. Modern English
2. Old English
3. Old Scandinavian
4. Middle English
5. Composite

The first recourse for choice of lexel is a Modern English identifier. This may be either a descendant or semantic equivalent of the Middle English form being tagged (e.g. $brother). Type 1 also includes combinations of Modern English identifiers that may not in fact be Modern English words in those combinations but whose elements are all Modern English (e.g. $unkinness ‘unnaturalness’).
In some cases there is no Modern English equivalent for our Middle English forms (e.g. flēmen ‘put to flight’, fraisten ‘inquire’). In these cases, we use approximate etyma as identifying labels, e.g. Old English $fle:man or Old Scandinavian $freista. We also utilise Old English or Old Scandinavian labels where the apparent Modern English equivalent would be ambiguous in a Middle English context. For example ModE ‘lie’, like ME līen, conflates two historically different words with different meanings: OE lēogan ‘tell a lie’ and OE licgan ‘lie down’. We therefore use the lexels $le:ogan and $licgan respectively.4 The Old English labels are normally in the West Saxon shape familiar from the standard dictionaries and grammars, even if these are not directly ancestral to any forms in the corpus. For instance, we sometimes use early West Saxon identifiers such as $cierran and $di:egel even though early West Saxon is not ancestral to late West Saxon, and late West Saxon itself is ancestral to only a small proportion of our corpus forms. Except where otherwise indicated, the Old Scandinavian labels (e.g. $-leikr ‘state’) are taken from comparable normalised Old Icelandic forms.
If a word of non-Scandinavian origin has no recorded or naturally reconstructable Old English form, and there is no unambiguous Modern English descendant, we use a Middle English identifier. This category includes loanwords from French, Middle Low German etc. Many of the identifiers in category 4 above will in fact be French loan words, but choosing one particular French etymon is often problematic given the range of both Anglo-French5 and Middle English spellings. In most of these cases therefore the MED headword has been chosen as the lexel.
Sometimes a lexel will be a composite label e.g. $Ya:that,6 which has an Old English element and a Modern English element. This is to facilitate comparison of each element with its equivalent simplex lexel.
Annotations to a lexel placed within braces before / and the following grammel refer to some semantic or functional distinction from otherwise identical unannotated or differently annotated lexels. For instance, $before{p}, $before{t} identify the lexel components of the tags $before/pr or $before/av (‘adverb’) when referring respectively to ‘place’ and ‘time’. Similarly, $be:am{l}and $be:am{t}refer respectively to reflexes of OE bēam meaning ‘beam of light’ and ‘tree, timber’. {*} indicates that the lexel label is an unattested word but is assumed to have been the origin of the forms it identifies, e.g. $gri:san{*} ‘terrify’, attested in Old English only with the prefix a-.
Separate lexels are given to prefixes, signalled as such by a trailing hyphen (e.g. $be-, $ge-, $un-), and to derivational suffixes, signalled as such by a leading hyphen (e.g. $-dom, $-hood, $-ly).


4.4 Grammels
The taxa invoked in the grammels are deliberately retrogressive. Our categories are ones that should be accessible regardless of a user’s theoretical orientation. They are deliberately ‘shallow’, i.e. they do not depend on potentially controversial hierarchical structure or on any theory of the binarity or otherwise of syntactic constituents. The tagging is lexico-grammatical in a purely surface sense: it taxonomises lexemes or grammatical items in their linear deployment. In the case of discontinuous structures, such as correlated negations, we also mark the co-membership in constructions of items split on the surface (see §§ 4.4.2.3, 4.4.5.1 and 4.4.6.3 below).
The taxonomic categories are essentially those of ‘traditional’ (Latinate) grammar. Almost all the categories defining the grammels (‘noun’, ‘object’, ‘conjunction’, ‘person’, ‘gender’) will be familiar to linguists or to any user with a modicum of traditional grammatical training. Conceptually we do not go much beyond the taxonomies utilised in 18th-century grammars: few of our taggings would be opaque to Bishop Lowth or Lindley Murray.
A grammel consists of everything in a tag appearing between / and _ , where / divides lexel from grammel and _ precedes a text form. In other words it is a complete grammatical characterisation of a form labelled by a particular lexel. Prefixes and derivational suffixes have grammels beginning /xp and /xs respectively followed by a specifier, indicating what part of speech they are affixed to.7


4.4.1. Nouns

4.4.1.1 Introduction
The basic label for a noun is n. We take nominative as the default case: singular noun subjects are not further specified. In other than subject position we label nouns by their argument functions, which may retain formal marking in the early Middle English period, requiring separate tagging for comparative purposes. The label nO designates ‘object’, which is always specified further as either Od (direct object) or Oi (indirect object). We use these terms in more or less the traditional way, except for the fact that functional indirect objects marked with a preposition are treated separately, along with other prepositional phrases. The reason we make this distinction is that nouns in this function overlap formally with those in adverbial and other prepositional phrases. A noun governed by an adposition is given the label n<pr (preposition) or n>pr (postposition):

$/RTApl_yE
$bewake/vpt23K2_BI-wAK+EDEN $be-/xp-v_BI- $/vpt23[K]K2_+EDEN
$/P23G_HERE
$orf/nOd_OREF
{=OREF underlined but no gloss given=}
{.'}
$beside/pr_BISIDES
$/T<pr_yE
{\}
$burh/n<pr_BUREHG
;_BETHLEEM

‘Who watched their flock beside the city Bethlehem’ (trhomBt.tag)8

Where a noun is used as a vocative, the annotation -voc is added to the basic n label:

$lord/n-voc_LOUerD
$say/vpt13_SEI+DE $/vpt13[V]_+DE
'_IOSEP
$/P12GN+C_yI
$swefn/n_SWEUEN
$be/vps13_IS
$e:aYe/aj_Ey
$to/im+C_TO
$tell/vi-m{rh}_TELL+E $/vi-m{rh}{rh}_+E

‘”Lord”, said Joseph, “thy dream is easy to interpret”’ (iacobt.tag)

Where a noun phrase alone serves an adverbial rather than an argument function, the annotation -av is added to the basic label:

$/P22N_zE
{\}
$all/ajpl_AL $/plaj_0
$fast/vps22_UEAST+Ed $/vps22_+Ed
{.}
$/T-av_yE
$sunday/n-av_SUNNE+DEI $day/n-av_+DEI

‘Ye all fast [on] the Sunday’ (corpart.tag)

4.4.1.2 Inflexions
Genitives and plurals are special cases in the tagging system. For the most part such forms are marked by suffixes, which need to be separately recoverable for comparative study and potential mapping. The basic label for genitive is nG, for plural npl. The suffixes respectively receive the tags $/Gn and $/pln. Plurals may be further characterised by labels like Od, <pr etc, in which case the separated suffixes would be labelled $/plnOd, $/pln<pr etc. In other words grammels can be built up incrementally to refine our characterisations. At this point we introduce an instance in which we appear to violate our own ‘surfacist’ principles. We explicitly tag zero elements in the case of genitive and plural. The reason for this is that zero, in contrast to non-zero marking, may be a spatiotemporal variable. Examples of non-zero and zero genitive tags are;

$father/nG_FADER+ES $/Gn_+ES
$father/nG_FADER $/Gn_0

+ indicates a formative boundary. A comparative table or a map of nominal singular genitive marking could thus display the relevant affixes independently of their base forms.
For athematic nouns that form their plurals by umlaut of the stem vowel, such as ‘man’ and ‘foot’, the plural endings are only hived off in those cases where the stem vowel has been analogically levelled from singular to plural and there is therefore a plural inflection comparable to other nouns. So contrasting examples would be tagged:

$man/npl_MEN
$man/npl_MANN+EN $/pln_+EN

There are some nouns whose plural inflections are irregular and/or may carry double plural inflection in e.g. <-ren> or <-nen>, for instance, ‘child’, ‘eye’, ‘ox’. The potentially detachable endings in these words are always left undetached: once endings are hived off from the parent word, they become anonymous when listed as suffixes in the text dictionaries. Non-comparable categories should therefore not be mixed with those that are strictly comparable.

4.4.1.3 Grammatical gender
We do not mark grammatical gender in nouns. Where grammatical gender still exists the key to its characterisation will always be carried by determiners or anaphoric pronouns, which we mark for semantic gender. See further discussion in §4.4.2 (a) below.

4.4.1.4 Noun as adjective
The label naj indicates a noun in adjectival function:

$/AN_*AN
$cloister/naj_CLOYSTer
$monk/n_*MONK
$love/vpt13_LOU+ED $/vpt13_+ED
$/P13OdM_HIM
$full{v}/av_FUL+
$well{w}/av{rh}_+WEL

‘A cloister monk loved him very dearly’ (edincmbt.tag)

It is arguable that the sequence ‘cloister monk’ could be interpreted as a compound rather than as a phrase with an attributive noun. Such judgements, however, even for modern English, may be fuzzy.9 The grammel here may serve as a pointer to precisely this fuzziness.

4.4.1.5 Compounds
The addition -k to the n-grammel marks the second element of a compound, e.g. the word ‘rood-tree’ as a prepositional object is tagged as follows:

$roodtree/n<pr_RODE-TRE $tree/n<pr-k_-TRE

All occurrences of the element ‘tree’ as second elements of compounds are thus available to be analytically treated along with forms of the simplex if desired.10

4.4.1.6 Titles and names
The addition of -t to the n-grammel marks a title, e.g. ‘saint’ or ‘king’ when used in immediate proximity to a proper name.

$saint/n-t_SEINTE
'_PAwEL

$of/pr_OF
$/T<pr_yE
$king/n<pr-t_KING
'_DAUIy

LAEME does not tag proper names. However, they are flagged for purposes of retrieval: personal names are preceded by '_ and place names by ;_.


4.4.2. Pronouns

4.4.2.1. Personal pronouns
(a) Basic categories
The basic label for a personal pronoun is P. The grammel is then built up in the following order: number, person, case, gender. Singular is designated by 1 and plural by 2; first person by 1, second person by 2 and third person by 3; for morphological reasons which do not apply to nouns, nominative is here specified by N. Other cases are characterised as with nouns. For instance, P11N represents ‘I’, P23Od represents ‘them’ as direct object. Two of these categories are found in the following example:

$more/ajn_*MO
$than/cj_yEN
$/P11N+V_ICH
$/P22Oi_OU
$tell/vi_TELL+EN $/vi_+EN
$can/vps11{rh}_CON

‘More than I can tell you’ (digby86mapt.tag)

For historical reasons we include genitives of personal pronouns as part of the pronoun paradigm, e.g. ‘our’ is labelled P21G. Structurally, however, as in modern English, the genitives of early Middle English personal pronouns function as a special subclass of ‘personal adjectives’, rather than as true members of the pronoun paradigm as they did in Old English (e.g. they no longer generally serve as verb objects). Because of this adjectival function, genitive personal pronouns require a double case marking, e.g.:

$&/cj_*AND
$with{w}/pr_WITH
$/P12G<pr+C_yI
$charter/n<pr_CHARTRE
$make/v-impK2_MAKE
$free/ajOd{rh}_FRE11

‘And with thy charter make free’ (havelokt.tag)

Here the label G<pr indicates a genitive personal pronoun functioning adjectivally after a preposition. Note the +C indicator. The spellings of ‘my’ and ‘thy’ may be responsive to the initial of the following word: before consonants there is a tendency for the final <n> to delete, before <h> or a vowel it tends to be retained. We therefore distinguish these using +C, +H and +V markers.12 This allows us to isolate the early stages of a process, which became virtually exceptionless by the late 16th century. Consider the following example:

$lord/n-voc_*LOUERD
$/P21N_WE
$be/vps21_AREN
$both/pn_BOyE
$/P12Gpnpl{rh}_yINE
{\}
$/P12GNpl+C_*yINE
$churl/npl_CHERL+ES $/pln_+ES
$/P12GNpl+H_yINE
$hi:gan/npl{rh}_HIN+E $/pln{rh}_+E

‘Lord we are both thine \ Thy churls and thy servants’ (havelokt.tag)

The second verse line illustrates the form of ‘thy’ retaining <n> before following initial <h>. The first verse line illustrates the marking (Gpn) assigned to the personal pronoun when it follows a copula. The extra marking, pn, was chosen to indicate a perceived ‘nominality’ where the essentially adjectival genitive appears in predicate rather than attributive position. This is another example of how a grammel may be incrementally constructed.

(b) Dual and other number categories
In early Middle English the dual still survives in some text languages. For operational convenience, and because it takes plural concord, we treat dual as a subcase of plural, marked under the relevant P21 and P22 grammels with D. E.g.:

$/P21GD_UNKER

‘Of us two’

The following text example illustrates the ‘polite plural’, i.e. ‘you’ used as singular. In such cases the number is marked with 0 because it would be misleading to mark it with either 1 or 2.

$lord/n-voc_LOUERD
$biddan/vps11+P_BIDD+I $/P11N_+I
$/P02Oi_YOU

‘Lord, I ask you’ (havelokt.tag)

Similarly, the ‘royal we’ is marked 01 rather than 21.

Semantic gender is marked with M = masculine, F = feminine, I = inanimate/non-human, e.g.:

$/AN_*A
$woman/n_wUmMON
$/RTA_yAT
$have/vps13_HAU+ED $/vps13_+ED
$lose/vpp_I+LOS+ED $ge-/xp-vpp_I+ $/vpp[S]K2_+ED
$/P13GF_HIRE
{\}
$needle/nOd_NELDE

‘A woman that has lost her (P13GF) needle’ (caiusart.tag).

(c) Grammatical gender
The grammatical gender systems in even the earliest Middle English texts are in transition to semantic ones. Marking forms for semantic gender regardless of morphology makes it possible to characterise the state of what grammatical gender system may still exist in a given text language. Consider the following example:

$bletsian/vppK2_*I+BLESC+ED $ge-/xp-vpp_*I+ $/vpp[S]K2_+ED
$be/vsjps13_BEO
$ever/av_EAUER
$/DisN_yOS
{\}
$hand/n_HOND
{.}
$for/cj_FOR
$/P13NI_HA
$have/vps13_HAU+ED $/vps13_+ED
$timber/vppK2_Y+TIMBR+ID $ge-/xp-vpp_Y+ $/vpp[R]K2_+ID
$/P11Oi_M>E>
{=E badly formed and repeated above=}
$/TplOd_yEO
$bliss/nplOd_BLISS+EN $/plnOd_+EN
{\}
$of/pr_OF
$heaven/n<pr_HEUENE

‘Blessed be ever this hand, for it (P13NI) has built for me the blisses of heaven’ (caiusart.tag).

Here the pronoun referring to ‘hand’ would appear from a formal historical point of view to be a feminine (for HA cf. OE hēo ‘she’); but the natural translation ‘it’ shows that it is being treated as a non-gendered noun, hence the marker I.13

(d) Reflexives
Grammels of personal pronouns used reflexively carry a final X, whether or not they are formally marked as reflexives:

$how/av_HOU
$/P23N_HI
$shall/vps23_SSOLLE
$/P23OdX_HAM-ZELUE $-self/xs-plOd_-ZELUE
$shrive/vi{rh}_SSRI\U+E $/vi{rh}_+E
{,}
$&/cj_AND
$make/viK2_MAK+I $/viK2_+I
$/P23OdX_HAM
$clean/ajplOd_KLEN+E $/plajOd_+E

‘How they shall shrive themselves and make themselves clean’ (ayenbitet.tag).

The marker X without further case specification is used when the reflexive pronoun is in apposition to a subject:

$/DatpnOd_*yET
$/P21G_OURE
$lord/n_LHORD
$/P13XM_HIM-ZELF $-self/xs_-ZELF
{,}
$/P21Oi_OUS
$forbid/vps13-ct_UOR+BYET $for-/xp-v_UOR+

‘Which our Lord himself us forbids’ (ayenbitet.tag).

When the reflexive pronoun itself serves as the subject the designation NX is adopted:

$as/RT_ALSE
{\}
$/P13NXM_HIM-SELF $-self/xs-N_-SELF
$say/vps13_SEI+d $/vps13[V]_+d

‘As himself says’ (trhomBt.tag).

We use the term ‘mental reflexive’, marked MX, for self-referential constructions of an essentially ‘middle voice’ type:

$when/cj_HUANNE
$/P13NM_HE
{,}
$/neg-v>=_NE
$bethink/vps13_BE+yENG+y $be-/xp-v_BE+ $/vps13[K]_+y
$/P13MXM_HIM
$not/neg-v<=_NAzT

‘When he bethinks him not’ (ayenbitet.tag).

4.4.2.2. Determiners
(a) Articles
The grammel for the definite article is T (‘the’); that for the indefinite article is A (‘a/an’). These function as premodifiers and will therefore always carry additional case or function marking:

$in{p}/pr_*IN
$/A<pr_A
$stone/n<pr_STON
{"cave"}
$still/av_STILLE
$/P13NM_HE
$licgan/vSpt13{rh}_LAI
{.}
$until/cj_TIL
$/P13NI_IT
$come/vSpt13_KAM
$/TN_dE
$3/qoaj_DRIDDE
$day/n{rh}_DAI

‘In a cave still he lay, till it came — the third day’ (bestiaryt.tag).

The label -ad added to the definite article grammel signifies that the preceding word ends in a dental stop; the added label -as that it ends in <s>. As the following examples show, either of these sequences may cause the following initial fricative to become a stop, though this effect after a preceding dental stop is more widespread:14

$&/cj_&
$smooth/vps13K2_SMEd+Ed $/vps13K2_+Ed
$/TOd_yE
$heart/nOd_HEORTE
{.}
$&/cj_&
$/TOd-ad_TE
$inwit/nOd_INwIT
$of/pr_OF
$sin/n<pr_SUNNE

‘... and smooths the heart and the conscience of sin’ (corpart.tag)15

$/P13NM_HE
$forlose/vSpt13_FOR+LEAS $for-/xp-v_FOR+
$/TOd-as_TE
$lorddom/nOd_LAUerD\+DOM $-dom/xs-nOd_+DOM

‘He lost the lordship’ (titushmt.tag).

There is one article-like item which is labelled differently: this is the reflex of old comparative instrumentals, e.g. ‘the better’, ‘the worse’. The determiner in these cases is labelled D-cpv:

$/D-cpv_yE
$long/av-cpv_LANG+ER $-er/xs-cpv_+ER
$that/cj_yAT
$/P11N+V_ICS
$/P13OdI_IT
$know/vps11{rh}_KNOWE
$/D-cpv_yE
$less/av_LASSE
$/P11N_ICS
$find/vps11_FINDE
$pris/nOd_PRIS
$therein{m}/av-k{rh}_yAR-ON $in{m}/pr-k{rh}_-ON

‘The longer I know it the less I find value in it’ (arundel248t.tag)

(b) Demonstratives
Like the articles, the demonstratives function most typically as modifiers. We group them under the pronominal heading largely for historical reasons and because they also occur in pronominal function. We divide the demonstratives into two classes: proximal and distal. The singular proximal demonstrative is labelled Dis, the plural Des. The grammels are designed as a mnemonic combining D for determiner and the syllable rhymes of ‘this’ and ‘these’, whether in Old or Modern English. For example:

$&/cj_&
{\}
$at/pr_ED
$/Dis<pr-ad_TIS
$word/n<pr_WORD

‘And at this word’ (corpart.tag).

As with the definite article, the added label -ad signifies that the preceding word ends in a dental. As this example shows, this sequence may cause the following initial fricative to become a stop. When a determiner functions as a pronoun, the label pn is added before the case marker:

$whoso/pn_HwA-SE
$witan{g}/vps13-ct_wIT
$/DespnOd_yEOSE
$well{w}/av_wEL
{.'}
$/P13NM_HE
{\}
$do/vps13_DE+d $/vps13[V]_+d
'_*SALOMONES
$bod/nOd_BODE

‘Whoso guards these well he does Solomon’s command’ (corpart.tag).

The distal demonstratives are labelled Dat and Dos (‘that’ and ‘those’) and are treated exactly the same as the proximal ones.


4.4.2.3 Relative markers
(a) Basic relative types
We group relative markers under the heading of pronouns for convenience. As in Modern English, early Middle English relative markers are not necessarily pronominal, though the pronominal forms may be considered prototypical and we therefore discuss them first.
Pronominal relatives are labelled RTA for animate human antecedent or RTI for inanimate, e.g.:

$forthy/cj_FOR-yI
$/P13NI_HIT
$be/vps13_IS
$right/aj_RIHT
$that/cj_y~
$/P21N_wE
$forlae:tan/vps21-apn_FOR+LET+EN $for-/xp-v_FOR+ $/vps21-apn_+EN
{*sj context*}
{.}
$&/cj_&
$forsake/vps21_FOR+SAK+EN $for-/xp-v_FOR+ $/vps21_+EN
{*sj context*}
$nightly/ajplOd_NIHT+LICH+E $-ly/xs-ajplOd_+LICH+E $/plajOd_+E
$deed/nplOd_DED+EN $/plnOd_+EN
{.}
$/RTIpl_yO
$be/vps23_BEN
{\}
$/TplN_yE
$work/npl_wERK+ES $/pln_+ES
$of/pr_OF
$Yy:sterness/n<pr_yIESTE>R>+NESSE $-ness/xs-n<pr_+NESSE
{=R interlined above=}

‘Therefore it is right that we forsake nightly deeds that are the works of darkness’ (trhomAt.tag).

Consider the following similar example:

$in{t}/pr_ON
$/Des<pr_yESSE
$3/qc<pr_yRE
$week/npl<pr_wUK+EN $/pln<pr_+EN
{.'}
{\}
$/RTIpl_yE
$be/vps23_BEN
$cleopian/vpp-plK2_CLEP+ED $/vpp[P]-plK2_+ED
$advent/n_ADUENT

‘in these three weeks that are called advent’ (trhomAt.tag).

Here the particle yE is marked as a notional plural for comparison with pronominal forms like yO in the previous example.

(b) Compound relatives
Compound relatives of the type ‘those who’ are given grammels reflecting their origin in the combination of determiner (or particle) plus pronoun (or particle), e.g.

$unwre:on/vsjps23_VN+wREON $un-/xp-v_VN+
$/P13OdI_HIT
$to/pr+H_TO
$/P23<prX_HAM-SEOL\UEN $-self/xs-pl<pr_-SEOL\UEN
$/DospnRTAplOi_yEO-yE
$/P13NI_HIT
$tolimpan/vps13_TO-LIMP+Ed $to-/xp-v_TO- $/vps13[P]_+Ed

‘Let them uncover it to themselves, those to whom it pertains’ (corpart.tag).

When compound relatives of the type e.g. ‘the one who’, are discontinuous the grammels carry the annotations >= and <= respectively to indicate the connection.16 For example:

$/Datpn>=_SE
$do/vps13_DE+d $/vps13[V]_+d
$/P13GM_HIS
$ae:ht/nOd_HEgHTE
$in{i}/pr_ON
$siker/aj<pr_SIKERE
$stead/n<pr_STEDE
{\}
$/RTA<=_y~
$send/vps13-ct_SENT
$/P13OdI_HI
$to/pr+H_TO
$heavenrich{k}/n<pr{rh}_HEUE+RICHE $rich{k}/n<pr-k{rh}_+RICHE


He puts his wealth in a secure place who sends it to heaven’s kingdom’ (digpmt.tag).

(c) Relative conjunctions and adverbs
Some relative markers are conjunctions or adverbs, e.g. ‘when’, ‘where’, ‘as’. These cases are given the appropriate lexel and a grammel beginning RT, e.g:

$&/cj_*AND
$forthy/av_FORyI
$at/pr_AT
$/Dat<pr_yAT
$blissful/aj<pr_BLIS+FUL $-ful/xs-aj<pr_+FUL
$time/n<pr{rh}_TYME
{\}
$when/RT_*QUEN
$holy/aj_HALI
$church/n_KIRC
$welcome/vps13K2_WELCUm+ES $/vps13[N]K2_+ES
$/P13OdM{rh}_HIME


‘And therefore at that blissful time when holy church welcomes him’ (edincmbt.tag).

This is a simple temporal relative. Adverbial relative grammels may however carry more information, e.g.:

$lipnen/v-imp_*LIPNE
$/P12N_yOU
$not/neg-v<_NOHUT
$to/pr_TO
$borrow/vn<prK2{rh}_BOREW+INGE $/vn<prK2{rh}_+INGE
{\}
$for/cj_*FOR
$/P12N_yOU
$shall/vps12_SHALT
$miss/vi_MISS+EN $/vi_+EN
$of/pr_OF
$many/aj<pr_MANI
$thing/n<pr{rh}_yINGE
{.}
{\}
$full{v}/av_*FUL
$le:of/aj_LEF
$when{that}/RTI_WEN
{"that"}
$/P12Oi_yE
$be/vsjpt13{rh}_WERE

‘Do not trust in borrowing, for thou shalt lack many a thing that would be very dear to thee’ (digby86hendingt.tag)

Here the annotation {that} is added to the lexel to indicate that, in spite of its form, the word does not have simple temporal reference but functions as a ‘that’ relative; and the annotation I (‘inanimate’) is added to RT to indicate this property of its antecedent. Consider also the following more complex example:

$/P13NM_HE
$make/vps13K2_MAK+Ed $/vps13[K]K2_+Ed
$/TOd_yE
$unbeliefful/ajOd_UN+BI+LE\+FULLE $un-/xp-aj_UN+ $be-/xp-aj_+BI+ $-ful/xs-ajOd_+FULLE
$man/nOd_MAN
$to/im+C_TO
$believe/vi-m_LEU+EN $/vi-m_+EN
$such/ajplOd>=_SwILCH+E $/plajOd_+E
$wile/nplOd_wIGEL+ES $/plnOd_+ES
{.'}
$as/RTIpl>pr<=_SwO
$/P11N+V_ICH
$ere/av_AR
$ymbe/pr<_EMBE
$speak/vSpt11_SPAC

‘He makes the unbelieving man to believe such wiles as I spoke about previously’ (trhomAt.tag).

Here the relative ‘as’ is connected to the antecedent ‘such’ by >= and <= as noted above. The grammel for ‘as’ also contains the information that the antecedent is plural and that it is in the scope of a postposition.

4.4.2.4. Definites, indefinites and numerals

(a) Definites and indefinites
The grammels for definites (‘the one’ ‘the other’) and the indefinite (‘another’) are built up from the grammels for the articles with the addition of aj or pn plus -def or -indef, e.g:

$&/cj_&
$say/vps13_SEI+S $/vps13[V]_+S
$evil/nOd_U\UEL
$by{re}/pr_BI
$other/pn-indef<pr_AN-OdER $/A<pr+V_AN-

‘And says evil concerning another’ (titusart.tag).

The indefinite is given a lexel ($other) followed in this case by the pn distinguisher. The first element is treated like a prefix and tagged separately as an ordinary indefinite article to facilitate comparison of its forms with those of the simplex indefinite article. The assignment of a lexel is mainly so that the definite and indefinite usages of ‘other’ may be compared with ‘other’ functioning as an adjective or as a simple pronoun; but it is also a judgement based on the denser semantic content of ‘other’ as opposed to the more prototypical pronouns.17 Consider also:

$ilk/Datpn_*yE-ILKE
$stop/vps13K2_STO>P>P+Ey $/vps13[P]K2_+Ey
$1/qcaj-defOd_yET-ON $/TOd_yET-
$ear/nOd_EARE
$mid{w}/pr_MID
$earth/n<pr_ERyE

‘That same [person] stops that one ear with earth’ (ayenbitet.tag).

The definite pronoun ‘that one’ is given a lexel in the form of a numeral ($1). The grammel here identifies it as a quantifier and cardinal (qc). See further below on numerals. In this example, ‘that one’ functions as an adjectival direct object, hence aj-defOd. Like ‘other’ the pronoun ‘ilk’ is given a lexel ($ilk). In tagging ‘ilk’ forms have been treated as complex versions of Dis, Des, Dat and Dos, in other words as emphatic deictics. The determiner element has not therefore been labelled separately but treated as a diacritic (cf. cliticised forms such as yILKE, yULKE).

‘Man’ used as an indefinite retains its nominal lexel for comparative purposes but because it frequently appears in reduced form, as the noun does not, it is differentiated by the grammel indef:

$for/cj_*FOR
$man/indef_ME
$/P13OdF_HI
$hold/vps13-ct_HALT
$lothly/ajOd_LOD+LICH $-ly/xs-ajOd_+LICH
$&/cj_&
$foul/ajOd{rh}_FULE

‘For one considers her lothly and foul’ (cotowlat.tag).

(b) Numerals
Numerals are considered a class of quantifiers and are separated according to cardinal and ordinal use. Lexels for numerals are the relevant arabic numbers. The basic grammels for cardinal and ordinal are qc and qo respectively, where q = quantifier. The rest of the grammel is built up in the usual way. The cardinal is taken to be a pure quantifier. Markings for case are given only for numbers 1 to 4, as in Old English. Cardinals functioning as pronouns are marked with pn. The ordinal is differentiated for adjectival (aj) and pronominal (pn) function and number and case indications are always given.

$/P11GN+C_*MI
$mouth/n_*MUy
$have/vps13_HAU+Ey $/vps13_+Ey
$2/qcG_TWEIRE
$kin/nplG_KUNN+E $/plnG_+E
$salve/nOd{rh}_SALUE

‘My mouth has salve of two kinds’ (jes29t.tag).

$in{t}/pr_*IN
$/T<pr_yE
$42/qoaj<pr_*TWO+AND+FOWERTIgyE $2/qc-k_*TWO+ $&/cj_+AND
$40/qoaj<pr-k_+FOWERTIgyE
$year/n<pr_gEARE
$of/pr_OF
$/P01G_VRE
$crown/vn<prF_CRUN+INGE $/vn<prF_+INGE

‘In the two and fortieth year of our crowning’ (huntproct.tag).

Note that the elements of compound numerals are tagged separately in addition to the full form. Non-final elements of compound ordinals are cardinals.


4.4.3. Adjectives

4.4.3.1 Basic categories
The basic label for adjective is aj. As with nouns, we take nominative as the default case: singular adjectives modifying subjects are not further specified. In other than subject position the labels given to adjectives match the labels of their head nouns.

$/TN_*yAT
$plea/n_PLAIT
$be/vpt13_wAS
$stiff/aj_STIF
$&/cj_&
$stark/aj_STARC
$&/cj_&
$strong/aj{rh}_STROnG

‘The debate was stiff and stark and strong’ (cotowlat.tag).

$/P13NM_*HE
$mo:t/vps13_MOxxT
$well{w}/av_wEL
$e:acan/vi_EK+ENN $/vi_+ENN
$many/ajOd_MANIg
$word/nOd_wORD

‘He might well add many a word’ (ormt.tag).

It is clear from these examples that attributive and predicate adjectives are tagged without differentiation. As with nouns, the marking of adjectival inflection is generally restricted to plural and genitive:

$with{w}/pr_WID
$word/npl<pr_WORD+ES $/pln<pr_+ES
$fair/ajpl<pr_FEIR+E $/plaj<pr_+E
$&/cj_AnT
$sweet/ajpl<pr{rh}_SWET+E $/plaj<pr{rh}_+E

‘...with fair and sweet words’ (tr323bt.tag).

$son/n_SUNE
{\}
$/TG_yES
$rich/ajG_RICH+E $/Gaj_+E
$king/nG_KING+ES $/Gn_+ES

‘Son of the powerful king’ (layamonAat.tag)

Note that the genitive is tagged whether or not there is an inflectional marking:

$/P13NM_*HE
$know/vpt13_KNOwN+ED $/vpt13[N]_+ED
$alone/aj_ONE
$each/ajG_ILC $/Gaj_0
$star/nG_STERR+E $/Gn_+E
$name/nOd{rh}_NAME

‘He alone knew each star’s name’ (genexodt.tag).

In this example the ‘ending’ is given as zero for comparative purposes, just as with nouns.
As with nouns, the addition of -k to an adjective grammel implies membership of a compound. The addition of -t implies that the adjective is part of a title.



4.4.3.2 The strong/weak distinction
The strong/weak adjective distinction does not survive intact into Middle English and we therefore do not attempt to tag it as such. Some adjective inflection does survive, but not all of it is salient or unambiguous enough to be tagged. There is for instance no doubt that an adjective marked with +ES is most likely to be a non-feminine genitive, as one marked in +NE is likely to be a masculine accusative. Where such endings survive we do tag them:

$/P13NM_HE
{\}
$have/vi_HABB+En $/vi_+En
$shall/vps13_SCAL
$good/ajOd_GOUD+NE $/Odaj_+NE
$rae:d/nOd{rh}_REaeD

‘He shall have good advice’ (layamonAbt.tag).

Outside these categories, most adjectives that have an apparent ending terminate in E. It is often difficult or impossible to tell if a final non-root E represents an inflection or is merely orthographic. This undecidability is impregnable in prose texts. Even in verse, the general variability of metre in early Middle English makes identification of final E as syllabic sufficiently problematic for us not to tag it.18

4.4.3.3 Adjectives in other part of speech functions
As with nouns, where an adjective phrase serves an adverbial rather than an argument function, the annotation -av is added to the basic label:

$mo:t/vpt11_MOST
$/P11N+H_IC
$have/vi_HABB+EN $/vi_+EN
$/A-av+V_AN
$a:nle:pig/aj-av_ALPI
$Yra:g/n-av_yRAgE
$some/ajOd_SUMME
$lis/nOd_LIS\SE

‘I must have for a single instant [i.e. single-instantly] some ease’ (lamhomA1t.tag).

The same convention is used for the survivals of old adverbial genitives:

$holy/ajpl_*HALI+E $/plaj_+E
$old/ajpl_AL\D+E $/plaj_+E
{~f18r~}
$ancra/npl_ANCR+ES $/pln_+ES
$/P13OdI_HIT
$may/vps23_MAHE
$do/vi_DO+N $/vi[V]_+N
$some/ajG-av_SUMM+ES $/Gaj-av_+ES
$way/nG-av_wEI+S $/Gn-av_+S

‘Holy old anchoresses may do it in some manner’ (corpart.tag).

The label ajn characterises an adjective in nominal function:

$/P13NM_HE
$ha:tan{c}/vSpt13_HEyTE
$all/ajplOd_ALL+E $/plajOd_+E
$/P13GMpl_HIS
$man/nplOd_MEN
$manship/nOd_MON\+SIPE $-ship/xs-nOd_+SIPE
$bewinnan/vi_BI-WINN+E $be-/xp-v_BI- $/vi_+E
$/TplOd_yE
$young/ajnplOd_zONG+E $/plajnOd_+E
$&/cj_AND
{\}
$/TplOd_yE
$old/ajnplOd{rh}_HOLD+E $/plajnOd{rh}_+E
$/TplOd_yE
$strong/ajnplOd_STROnG+E $/plajnOd_+E
$&/cj_AND
$/TplOd_yE
$bold/ajnplOd{rh}_BOL\D+E $/plajnOd{rh}_+E


‘He called all his men to win honour — the young and the old the strong and the bold’ (layamonBOt.tag).

4.4.3.4 Comparison
Adjectival comparison is indicated by the labels -cpv (comparative) and -sup (superlative):

$/RTApl_yAT
$say/vpt23_SEI+DEN $/vpt23[V]_+DEN
$/P23N_HE
$be/vpt23_WEREN
$fair/ajpl-cpv_WEIR+ORE $-er/xs-cpv_+ORE
$than/cj_yEn
$god/n_GOT

‘Who said they were fairer than God’ (tr323at.tag).

$biddan/vsjps21_*BIDDE+
$/P21N_+WE
$/P21G_URE
$lady/nOi_LEUEDYE
$sweet/ajOi-sup_SUET+IS $-est/xs-sup_+IS
$all/ajplG_AL+RE+ $/plajG_+RE+
$thing/nplG{rh}_+yING+E $/plnG{rh}_+E

‘Let us pray to Our Lady, sweetest of all things’ (tr323at.tag).

For the grammels aj-def and aj-indef see §4.4.2.4 above and the following examples:

$by{p}/pr_*BI
$1/qcaj-def<pr_yE-ON $/T<pr_yE
$side/n<pr_SIDE
$/P13NM_HE
$see/vSpt13_SEI
$/AOd+V_AN
$isle/nOd_ILE

‘On the one side he saw an isle’ (corp145selt.tag).

$/P13NF_HO
$see/vSpt13_SEI
$other/aj-indefOd_AN+OyER $/AOd+V_AN+
$devil/nOd_DEUEL
$therein{p}/av-k_yER-INNE $in{p}/pr-k_-INNE
$Ya:/av{rh}_yO

‘She saw another devil therein then’ (tr323bt.tag).


For verbal adjectives see §4.4.4.8 (c) and (d) and §4.4.4.9.1 (b) below.


4.4.4. Verbs
The basic label for a verb is v. Verb features tagged include tense, number and person. The marking of mood is confined to forms of the subjunctive where it differs from the indicative. We also mark past and present participles, verbal nouns and infinitives. The normal procedure is to hive off affixes for the purposes of separate retrieval and comparison — prefixes and the suffixes marking present indicative 2nd and 3rd singular and all plurals, all past tenses, past and present participles, infinitives, plural imperatives and verbal nouns (cf. LALME item numbers 57–64). These are the elements that we consider to have potential geographical or temporal significance.19

4.4.4.1 Tense, number and person: introduction
We assume a two-tense system: vps = verb present, vpt = verb past/preterite. Grammels are built up in the same way as they are for other categories, as concatenations of features. The order we have chosen is number followed by person: e.g. $/vps13 = verb present singular third person; $/vpt21 = verb past plural first person:20


$/TN_y~
$3/qoaj_yRIDDE
$thing/n_yING
$when/RT_HwANNE
$/P21G_URE
$dryhten/n_DRIHTEN
{\}
$/P21Od_US
$biddan/vps13-ct_BIT
$turn/vi_TurN+EN $/vi_+EN
$to/pr+H_TO
$/P13<prM_HIM
{.'}
$be/vps13_IS
$wo:p/n_wOP
$/RTIOd_yE
$/P21N_wE
$for/pr_FOR
$/P21G_URE
$sin/npl<pr_SYNN+ES $/pln<pr_+ES
$weep/vps21_wEP+Ed $/vps21_+Ed
{.}
$as/RT_ALSE
$/TN_dE
$holy/aj_HOLIE
{\}
$prophet/n_ProPHetE
$say/vps13_SEI+d $/vps13[V]_+d
{.}
{(*LACRIMIS MEIS STr^ATUm MEUm RIGABO .(}

‘The third thing when our Lord bids us turn to him, is weeping, which we weep for our sins, as the holy prophet saith: I water my couch with my tears’ (trhomBt.tag).

In this example, note the tagging of the following forms:
(a) wEP+Ed, 1st person plural present indicative — $weep/vps21 with a separate tag $/vps21 for the separated inflection +Ed;

(b) BIT, 3rd person singular present indicative, where the ‘ending’ is assimilated to the dental stem giving a syncopated form from which no suffix can be separated on the surface. The grammel in such cases (confined to 3rd person singular present indicative) is marked -ct (contracted);

(c) SEI+d 3rd person singular present indicative, where the grammel of the detached ending is marked with a [V] to indicate that the verb it belongs to has a stem-final vowel. As can be seen here, this has implications for the shape of the inflection, which is syncopated.

4.4.4.2 Marked stem finals
The following numbers, tenses and derived forms are marked for stem final vowels because of potential syncope: vps12, vps13, vps2 (all present indicative plurals) vpt (all weak past tense forms). Also (see §§4.4.5.7–4.4.4.9. below) v-imp22 (imperative plurals), vpp (weak past participles), vSpp (strong past participles — S = strong), vSpt (all strong past tenses), vi (infinitive), vn (verbal noun = gerund), vpsp (present participles).

In 2nd and 3rd person singular indicative a number of other stem finals may also cause syncope in some text languages (cf. also past indicative and weak past participles §4.4.4.5 and §4.4.4.8 (d) below). In these cases we therefore mark the detached ending with a label, indicating the nature of the preceding stem final element as follows:

(a) nasals marked [N] — cf. CUM+y and CUM+Ey
(b) [r] marked [R] — cf. BER+ST and BER+EST
(c) [l] marked [L] — cf. FAL+y and FALL+ES
(d) [x]21 marked [H] —  cf. BOUz+y and BUG+Ed
(e) velar stops marked [K] — cf. BREC+ST and BREK+EST, BRING+Ey and BRENG+y
(f) sibilants marked [S] — cf. CUS+T and CUSS+ES
(g) [p] marked [P] — cf. LHAP+y and LEP+Ed
(h) <f> marked [F] — cf. B+LEF+y and BI-LEAF+EST. Compare also BE-LEU+y and BI+LEU+Ed (not marked [F])22
(i) <w, ƿ> are marked [W] — cf. BLAW+S and BLOUw+ET.23
(j) [j] marked [J] — SEg^g+y.24
(k) [t, d] marked [D] —  cf. HALD+ST and HOLD+EST. This marker is confined in the present tense to vps12 forms as dental stems in the 3rd person are either assimilated, providing no detachable ending (marked /vps13-ct), or are unsyncopated and as such are fully comparable to other unmarked unsyncopated endings.

The stem-finals listed in (a)–(k) above are the only ones whose detached endings are given descriptive markers. The endings are marked in all cases, whether or not the stem-final element has in fact triggered syncope. The default is lack of marking, whether or not, in exceptional cases, the detached ending is syncopated.
One verb, the survival of OE hātan in the sense ‘be called, be named’, has a unique set of specifiers. The passive form hatte survives into Middle English. In the 3sg present indicative, this form is in contrast to both unsyncopated forms, e.g. ATT+Ed, HAT+Ed, HOT+Ey, (unmarked by a specifier) and to syncopated forms, where the ending has been assimilated into the dental stem, e.g. HAT, HATE (marked -ct). The active and passive forms are already showing signs of blending, active verb spellings being frequently used also for passive senses. Where it occurs, the passive form is labelled $ha:tan{n}/vps13-passform, where {n} at the end of the lexel indicates that the verb means ‘is named’ rather than ‘command’ (labelled {c}) or ‘promise’ (labelled {p}):

$/P23N_HI
$resemble/vps23F-apn_RESEMBL+Ey $/vps23F-apn_+Ey
$/AOd+V_AN
$adder/nOd_EDDRE
{,}
$/RTI_yET
$ha:tan{n}/vps13-passform_HATTE
$siren/n_SERAYN

‘They resemble a serpent that is named Siren’ (ayenbitet.tag).

In early Middle English the gradual process of subsitution of the past tense forms of hātan for those of the present had already begun. For the 3sg present indicative these forms (either of the het-type or the hight-type) also come into contrast with the other present tense forms listed above. These past tense presents are labelled $ha:tan{n}/vps13-ptform, or $ha:tan{c}/vps13-ptform:

$for/pr_FOR
$/A<pr_ON
$sweet/aj<pr_SUETE
$wife/n<pr{rh}_WIF
{\}
$/RTA_*yAT
$ha:tan{n}/vps13-ptform_HEIzTTE
'_MARGERI

‘For a sweet woman who is named Margery’ (digby86siritht.tag).

$&/cj_&
$do/vsjps21-bpn_DO
$/P21N_wEx
$/RTIOd_y~
$/P13NM_HEx
$/P21Oi_US
$ha:tan{c}/vps13-ptform_HExT

‘.. and let us do what he commands us’ (egpm1t.tag).

4.4.4.3 The Northern Present Tense Rule
In northern and northerly text languages, immediate proximity of the verb to a personal pronoun may cause reduction or loss of the plural present indicative suffix. This is known as the Northern Present Tense Rule. For the purposes of retrieval and comparison, in all texts (whether northerly or not) the specifiers -bpn (before pronoun) and -apn (after pronoun) are added to the grammels of relevant plural verbs:

$/P23N_*yAI
$kindle/vps23-apn_KINDEL $/vps23-apn_0
$barat/nOd_BARET
$with{w}/pr_WIT
$backbite/vn<pr{rh}_BAC+BIT+ING $bite/vn<pr-k{rh}_+BIT+ING $/vn<pr{rh}_+ING
{\}
$&/cj_*AND
$slokna/vps23_SLOKEN+ES $/vps23_+ES
$/P13OdI_IT
$with{w}/pr_WIT
$/P23G_yAIR
$glo:ther/vn<pr{rh}_GLOyER+InG $/vn<pr{rh}_+InG

‘They kindle strife with backbiting and quench it with their flattery’ (edincmbt.tag).

Contrast endingless KINDEL, immediately following the personal pronoun, with the form SLOKEN+ES with normal northern +ES suffix.25

4.4.4.4 First person singular present indicative
Earlier in this exposition (§4.4.3.2), a conceptual problem arose, whose solution may have seemed partly arbitrary. When is a particular graphical object not analysed as representing a morph? The problem arose in the tagging of adjectival inflections: there were classes of final E that we decided not to tag and some that we did. This problem arises again with the first person singular present indicative. Unlike those of the second and third person singulars, the endings of 1st person singular forms are not normally hived off and given separate grammels because they yield no significant spelling variants to compare; and because, with some important exceptions to be detailed below, it is undecidable whether the final E forms represent anything phonological.

$for/cj_FOR
$now/av_NOU
$/P11N+V_ICH
$see/vps11_I+SEO $ge-/xp-vps11_I+
{=Final O not erased=}
$/P11GOd+V_MYn
$end/nOd{rh}_ENDE
{\}
$with{w}/pr_*WYy
$/P22<pr_zOU
$/P11N+H_ICH
$have/vps11_HABBE
$in{c}/pr_IN
$joy/n<pr_IOIE
$be/vSpp_I+BE[O] $ge-/xp-vpp_I+

‘ ... for now I see my end \ I have been with you in joy ... ’ (corp145selt.tag).

In LAEME we are mainly concerned with formal variation. Variation in the 1st person singular present indicative suffix of verbs is, in general, confined to the presence or absence of final E. In regular verbs final E occurs after consonantal stems (HABBE) and zero after vocalic stems (I+SEO). There are for the most part no regional or temporal implications.26 In some northern and northeast Midland texts, however, there is very sporadic evidence of the analogical extension of 3rd person singular endings across the paradigm. When these occur in the 1st person they are tagged separately:

$/P11N_HY
$forsake/vps11_FOR-SAK+IS $for-/xp-v_FOR- $/vps11_+IS
{=sic=}
$/TOd_YE
$world/nOd_WERLD

‘I forsake the world’ (merton248t.tag).

There are some other 1st person singular present indicative endings that we consider worth marking separately also: the reflexes of OE class II weak verbs, and some potentially comparable types. Preservation of thematic -i- in the present systems of these verbs is of great historical and geographical interest. The grammels of class II weak verbs are given the additional specifier K2.27 The K2 specifier is carried also by the grammels of all separated suffixes in these verbs, in order to differentiate them from the suffixes of originally non-class II verbs. Very often original thematic -I- is preserved in the first person singular present indicative, even where it has been lost elsewhere in the verb. In these circumstances it is clear that separate marking is desirable:

$for/cj_*FOR
$/P11N_ICH
$love/vps11K2_LOU+IE $/vps11K2_+IE
$christian/ajplOd_Cr^ISTEN $/plajOd_0
$man/nplOd_MEN
{.}
$&/cj_AND
$nourish/vps11Fir_NORISS+I $/vps11Fir_+I
$/P23Od_zAM
$also/av{rh}_ALSO

‘For I love Christian men and nourish them also’ (corp145selt.tag).

The verb love (OE lufian) here retains the +IE suffix in vps11. The grammels of both word and detached suffix are annotated with K2. Notice also NORISS+I. The +I suffix of this Old French verb (norrir) is potentially comparable with the native class II weak type. We also find +I(E) suffixes in OF verbs of the -ier conjugation. The annotation F (French) is added to the grammels of verbs that are French loans. Those of the -er and -re conjugations are given no further specification and the vps11 endings are not normally tagged separately. The -ir and -ier conjugations are marked Fir and Fier respectively and the suffixes of these verbs, like those of OE class II, are always tagged separately for comparison, whether or not thematic -i- is present (see further §4.8.8.8).28
The history of OE class II weak verbs includes not only the survival in some cases of thematic -I-, but also the adoption of -I- into verbs that were not class II weak in Old English. We therefore also tag separately all vps11 suffixes that have -I(E)/-Y endings. These class-shift examples are easily retrievable because they lack any further specifier in the grammel:

$/P11N_*IH
$acknowledge/vps11_KNOV+LECH+Y $-lae:can/xs-vps11_+LECH+Y $/vps11_+Y
$to/pr_TO
$god/n<pr_GOD

‘I confess to God’ (emmanuel27t.tag).

4.4.4.5 Weak past tense
The weak past tense is labelled vpt. Endings are hived off in the same way as for present tense indicative, though in this case for all numbers and persons. The exception is 3rd person singular forms with dental stems where the ending has been assimilated.

$/P13NM_HE
$wealdan/vpt13_WELDE
$/TOd_yAT
$rich{k}/nOd_RICHE
$here/av_HaeR
$all/ajpl_*AL $/plaj_0
{\}
$/P23N_HIT
$/P13OdM_HINE
$love/vpt23K2_LUU+EDE $/vpt23K2_+EDE
$/RTApl_yAT
$live/vpt23_LIU+EDE $/vpt23_+EDE
{=MS originally LIIUEDE or LUUEDE with second minim subpuncted=}
$in{p}/pr_IN
$land/n<pr_LOn\DE

‘He ruled the kingdom here. They all loved him that lived in the land’ (layamonAbt.tag).

Notice LIU+EDE and LUU+EDE both 3rd person plural and with formally identical suffixes, LUU+EDE, however, being marked with K2 as an original OE class II verb. Notice WELDE where detaching of the past tense suffix is impossible because of coalescence with the stem final.

As with vps12 and vps13, detached suffixes are marked with specifiers for stem-final consonants where these have a potential effect on their form:

$but/cj_*BOT
$for/cj_FOR
$/P11N_*I
$love/vpt11K2_LUF+ED $/vpt11[F]K2_+ED
$well{w}/av_WEL
$/P21G_OUR
$lady/nOd{rh}_LEFDYE
{\}
$while/cj_*QUIL
$/P11N_*I
$live/vpt11_LIF+D $/vpt11[F]_+D
$/P11N+H_*IC
$have/vpt11_HAF+D $/vpt11[F]_+D
$forthy/av{rh}_FOR-yIE
{\}
$full{v}/av_*FUL
$good/ajOd_GOD
$help/nOd_HELP

‘But because I loved Our Lady well, while I lived I had therefore very good help’ (edincmbt.tag).

Notice LUU+EDE, LIF+D and HAF+D all with stem final F. LUU+EDE, marked K2, unsurprisingly retains the thematic vowel (< OE -o-) in its suffix. In LIF+D and HAF+D the form of the suffix is apparently affected by preceding F, presumably here voiceless, and are vowelless.29 Other stem finals may cause devoicing as in Old English:

$/TN_yE
$holy/aj_HALI
$man/n_MON
$in{p}/pr_*IN
{('*VITAS \ PATRUM' .(}
$kiss/vpt13_CUS+TE $/vpt13[S]_+TE
$&/cj_&
$bletsian/vpt13K2_BLESC+EDE $/vpt13[S]K2_+EDE
$/TG_yE
$other/pnG_OyER+ES $/Gpn_+ES
$hand/nOd_HOND

‘The holy man in Vitas Patrum kissed and blessed the other’s hand’ (caiusart.tag).

In CUS+TE the detached suffix is marked with the specifier [S] so that this type may compared with those in similar phonetic contexts and differentiated from those in contrasting ones. In BLESC+EDE the class II thematic vowel is again preserved. The specifier [S] is used because we assume that +SC+ here stands for either [s] or [ts]. In the following example, the stem final dental has coalesced with the suffix, which cannot therefore be hived off in the usual way. But the plural ending, after the coalesced dental, is of interest and is therefore detached but marked [D]:30

$as/av>=_AS
$often/av_OFTE
$as/cj<=_AS
$/P23N_HA
$guilt/vpt23_GULT+EN $/vpt23[D]_+EN

‘... as often as they transgressed...’ (caiusart.tag).

The same precedure is followed with stem final dentals in the second person singular:

$/RTA_y^T
$sce:adan/vpt12_SSADD+EST $/vpt12[D]_+EST
$for/pr_FOR+
$/P11<pr_+ME
$/P12GOd+C_yI
$blood/nOd{rh}_BLOD

‘...who sheddest for me thy blood...’ (corp145selt.tag)

4.4.4.6 Subjunctive
The basic label for subjunctive is sj. Grammels containing it are built up in the order verb, subjunctive, tense, number, person. Typical grammels would be vsjps12, vsjpt23. The subjunctive as a formal category is already recessive in early Middle English, i.e. commonly undifferentiated from the indicative. We therefore confine the sj label to verbs that are formally differentiated from the corresponding indicatives:

though/cj_yAH
$/P13NF_HA
$li:chamli:ce/av_LICOM+LICHE $-ly/xs-av_+LICHE
$wunian/vsjps13K2_wUNIE
{\}
$upon/pr_UP-ON
$earth/n<pr_EORdE

‘... though she bodily dwell upon earth’ (bod34t.tag).

Here the form wUNIE lacks the final dental fricative, which marks the indicative. Plural subjunctives can be differentiated from indicatives only in those text languages that do not have ‘Midland -n’ in the indicative plural. In the following example the relevant text language consistently retains the dental fricative ending for plural present indicative:

$oYYaet/cj_A-yET
$/P23N_HA
$arise/vsjps23_A+RISEN $a-/xp-v_A+
$through/pr_yURH
$behre:owsian/vn<prK2_BI+REOwS+UNGE $be-/xp-v_BI+ $/vn<prK2_+UNG

‘... until they rise up through repentence’ (bod34t.tag).

Here A+RISEN is marked as subjunctive because it is formally differentiated from the indicative. As with the wUNIE example above, the subjunctive mood has apparently been triggered by the preceding conjunction. In early Middle English conjunctions that trigger subjunctive include ‘though’, ‘unless’, ‘until’ but, interestingly, not ‘if’, which normally seems to be followed by the indicative.

Consider the following example:

$well{w}/av_wEL
$be/vsjpt13_wERE
$/P23Oi_HAM
$be/vpt23_wEREN
{*sj context*}
$/P23N_HA
$on{t}/pr_ON
$/P23G_HARE
{\}
$bridela:c/nG_BRUD+LAK+ES $-la:c/xs-nG_+LAK+ES $/Gn_+ES
$day/n<pr_DEI
$bear/vSpp-pl_I+BOR+EN $ge-/xp-vpp_I+ $/vSpp-pl_+EN
$to/im+C_TO
$bebury/vi-m_BI+BURI+EN $be-/xp-v_BI+ $/vi[V]-m_+EN

‘It would be well for them were they to be borne away to be buried on their wedding day’ (bod34t.tag).

wERE is marked vsjpt13 because it evidently subjunctive: the indicative would be a form of ‘was’. The following plural wEREN is, however, formally impossible to differentiate from the indicative. We therefore tag it as indicative but indicate in commentary that the context suggests subjunctive. This {*sj context*} comment is also added to formally ambiguous verbs that follow the subjunctive-triggering conjunctions mentioned above. Differentiation of past subjunctives in strong verbs31 is possible for 1st and 3rd persons singular only by observing the stem vocalism and/or the presence of final <e>:

$if/cj_*zIF
$1/qcpn_ON
$of/pr_OF
$/P22G_OUWER
$ass/npl<pr_ASS+ENE $/pln<pr_+ENE
$in{i}/pr_IN
$/A<pr_A
$pit/n<pr_PUT
$fall/vsjpt13_FULLE
$today/av{rh}_TO-DAY

‘If one or your asses should fall into a pit today... ‘ (laud108at.tag)

The expected vocalism for the past indicative would be <e> or <a>. Contrast the following example where the stem vocalism in the 2nd person singular past tense is the same for indicative and subjunctive and the form therefore has been marked indicative but a note given that the context suggests subjunctive:

$how/av_*HOU
$may/vpt13_MIzTE
$/P13NF_HE[O]
$see/vi_I+SE[O] $ge-/xp-vi_I+ $/vi[V]_0
$cwellan/vi_QUELL+E $/vi_+E
$/P13GF_HURE
$child/nOd_CHILD
{.}
$that/cj_y^T
$/P13GF_HURe
$heart/n_HERTE
$/neg-v_NE+
$break/vSpt12_+BREKE
{*sj context*}
$in{m}/pr_A+
$2/qcpn<pr{rh}_+TO

‘How could she see her child killed, that her heart should not break in two’ (corp145selt.tag)

Sometimes only the presence of final <e> differentiates singular strong past tense subjunctive from indicative. Since some text languages show final <e> also in indicative singular, decisions about whether to tag the form as subjunctive or to mark it as indicative but in a subjunctive context have to be made on a case by case basis:

$that/cj_*yAT
$ilk/DosOd_yULK+E $/DosOd_+E
$gerd/nplOd_zERD+EN $/plnOd_+EN
$3/qcOd_yRE[O]
$fetch/vpt13_VETTE
{*sj context*}
{.}
$&/cj_AND
$with{w}/pr_WIy
$/P13<prXM_HIM
$niman/vsjpt13{rh}_NOME
{\}
$9/qc_*NIE
$day/npl-av_DAW+ES $/pln-av_+ES
$/P13NM_HE
$be/vpt13_WAS
$thitherward/av_yUDERWARD
{.}
$ere/cj_AR
$/P13NM_HE
$thither/av_yUDER
$come/vsjpt13{rh}_COME
{\}
$with{w}/pr_*WIy
$great/aj<pr_GRET
$honour/n<pr_HONur
$/P13NM_HE
$niman/vSpt13_NOM
$/P23Od_IS
$up/av_UP

‘... that he should fetch those same three sticks and should take them with him. He was nine days on the way before he might come thither; with great reverence he took them up...’ (corp145selt.tag).

Here the form of the weak verb FETTE is the same for indicative and subjunctive. But observe the contrast between NOME and COME, where final <e> indicates subjunctive, and the indicative form NOM.
Hortative subjunctives immediately preceding the pronoun ‘we’ (i.e. the type translated ‘let us’ in Modern English) appear to have reduced endings in many text languages. Compare:

$oYYaet/cj_AyET
$/P21N_wE
$speak/vsjps21_SPEOKEN

‘... until we may speak ...’ (corpart.tag).

with:

$come/vsjps21-bpn_*CUME
$/P21N_wE
$now/av_NU
$eft/av_EFT
$again/av_A\zEIN
$&/cj_&
$speak/vsjps21_SPEOKEN
$of/pr_OF
$all/ajnpl<pr_ALL+E $/plajn<pr_+E
$gemae:ne/av_I+MEANE $ge-/xp-av_I+

‘Let us proceed now again and speak of them all in common’ (corpart.tag).

Reduced *CUME and unreduced SPEOKEN seem here to follow a pattern similar to that of the ‘Northern Present Tense Rule’ for plural indicatives (see §4.4.4.3 above), even though the text in question (Ancrene Wisse), and many others that behave the same way, are not northerly in origin. As the above example shows, we have marked these hortative /vsjps21 forms with the specifier -bpn.

4.4.4.7 Imperative
This term imperative is here restricted to ‘true’ imperatives, i.e. not hortatory or 3rd person imperatives, which are formally subjunctive and marked as such. The basic label for imperative is v-imp. For singular imperatives no further specification is given. Plurals are marked with additional 22 and their endings are hived off for separate tagging:

$help/v-imp_*HELP
$/P21Od_US
$lady/n-voc_LEUE>DI>
{=DI interlined above=}
$/T-av_yE
$stund/n-av{rh}_STUNDE
{\}
$when/RT_WEN
$/P13NM_HE
$show/vps13K2_SAU+IT $/vps13[V]K2_+IT
$/P13GM_IS
$te:ona/nOd{rh}_TEnNE
{\}
$go/v-imp22_*GO+D $/v-imp22[V]_+D
$/P22-voc_yE
$awaergan/vpp-ajn-vocpl_A+WARI+EDE $a-/xp-v_A+ $/vpp[V]-ajn-vocpl_+EDE
$into/pr_In-TO
$woe/n<pr{rh}_WO

‘Help us, Lady, at the time when he shows his anger: “Go, ye cursed, into woe”’ (tr323at.tag).

Where the personal pronoun immediately follows plural imperatives, the verb often loses its suffix. When the personal pronoun is retained, the plural imperative is therefore marked -bpn, like plural indicatives and hortative subjunctives in immediate proximity to personal pronouns (see §§ 4.4.4.3 and 4.4.4.6).32
Singular imperatives are normally endingless. In exceptional cases an imperative in a singular context is given an ending. In these few cases the ending is apparently extended from plural indicative or subjunctive. In these examples we give the separated ending the grammel v-imp12.

$tell/v-imp_TELL+EN $/v-imp12_+EN
$all/ajnOd_AL
{.}
$sir/n-voc_*SIRE
$/P11N+H_ICH
$have/vps11_HAB\BE
$/DispnOd_yIS
$thus/av_yUS
$often/av_OFTE
$do/vSpp_I+DO+N $ge-/xp-vpp_I+ $/vSpp[V]_+N

‘“Tell [speaking to a nun] all”. “Sir, I have done this often”’ (caiusart.tag).

4.4.4.8 Non finite categories
There is a number of derived categories which are traditionally associated with the verb paradigm in the western grammatical tradition. These are the verbal nouns — the infinitive and the gerund — and the present and past participles.

(a) Infinitive
The basic label for infinitives is vi. The OE inflexions of the infinite — -an, -ian and -(e)nne — were already beginning to be reduced in later Old English. In early Middle English there are some survivals of original -a- in the infinitive ending but mostly the vowel is ­-e-, with or without surviving final -n. For comparison, therefore, verbal infinitives of regular verbs always have their endings hived off and tagged separately, even if this ending is lost. In these cases the ending is realised in the tagging as zero:

$so/av_SO
$help/vsjps13_HELP
$/P11Od_ME
$christ/n_Cr^IST
$if/cj_AND
$/P11N_HY
$may/vps11_MAY
$spe:dan/vi{rh}_SPED+E $/vi{rh}_+E
{\}
$richly/av_*RIC+LIC $-ly/xs-av_+LIC
$shall/vps12+P_SAL+TU $/P12N_+TU
$have/vi_HAF $/vi_0
$/P12GOd+C_YI
$meed/nOd{rh}_MEDE

‘So help me Christ, if I may succeed, richly thou shalt have thy reward’ (clericot.tag).

The retention, reduction or loss of infinitive endings in early Middle English may be affected by a number of factors. As with the finite categories, the basic vi label can carry additions reflecting these factors as follows:
(i) K2 for OE class II verbs where survival of thematic <i> may be expected;
(ii) K2[1] for OE class I verbs with thematic <i>;
(iii) F for verbs that are French loans of the -er and -re conjugations;
(iv) Fier and Fir for French loans of the -ir and -ier conjugations respectively;
(v) [V] for verbs with stem final vowel which may trigger syncopation. This is the only stem final that is marked for infinitives, as it is the only type that consistently has the power to affect the form of the ending.

In the following example, note the survival of thematic <i> in both ‘wake’ (< OE class II weak wacian) and ‘wear’ (< OE class I weak werian). The additions to the grammel ensure that these infinitive endings can be sorted separately and compared with others from verbs of the same class that may not retain thematic <i>:

$/Datpn_y~E
$teach/vps13_TECH+Ed $/vps13_+Ed
$all/av_AL
$how/av_HU
$man/indef_ME
$shall/vps13_SCHAL
$bear/vi_BER+EN $/vi_+EN
$/P13OdXM_HIM
$without{os}/av_wId-VTEN
{\}
$how/av_HU
$eat/vi_ET+EN $/vi_+EN
$drink/vi_DRINCK+EN $/vi_+EN
$wear/viK2[1]_wER+IEN $/viK2[1]_+IEN
$licgan/vi_LIGG+EN $/vi_+EN
$sleep/vi_SLEP+EN $/vi_+EN
$wake/viK2_wAK+IEN $/viK2_+IEN

‘That reaches how one should in all respects bear oneself outwardly: how to eat, drink, dress, rest, sleep and wake’ (neroart.tag).

Compare the labels on the infinitives in the example below:

$/P12N_yOU
$/neg-v(>but>neithernor)_NE
$shall/vps12_SSELT
$have/vi_HABB+E $/vi_+E
$god/nOd_GOD
{,}
{\}
$bu:tan{e}/av_BOTE
$/P11Od_ME
{.}
$neither/cj>=_NE
$worship/vi_WOR+SSIP+IE $-ship/xs-vi_+SSIP+IE $/vi_+IE
{,}
$nor/cj<=_NE
$serve/viFir_SERU+I $/viFir_+I
{.}
$&/cj_*AND
$/P12N_yOU
$/neg-v(>but)_NE
$shall/vps12_SSELT
$do/vi_DO $/vi[V]_0
$/P12GOd+H_yINE
$hope/nOd_HOPE
{,}
$bu:tan{o}/av_BOTE
$in{p}/pr_INE
{\}
$/P11<pr_ME

‘Thou shalt not have a god except me, either to worship or to serve; and thou shalt put thy trust in me only’ (ayenbitet.tag).

Here HABBE has the expected +E ending, reduced from OE ­-an. The French loan SERUI from servir retains thematic +I in its ending. But the +IE ending has been extended from its use in OE class II and French -i(e)r loans to WOR+SSIP+IE, a verb with no direct OE antecedant, but which was developed from the noun. Most verbs of this type have regular -e(n) endings in Middle English. Leaving the vi grammel unmarked by any further label ensures that this ending is grouped for comparision with the expected regular endings of this type, and the unusual extension is thereby salient. With DO, the grammel of the ‘hived off’ zero ending is given the added label [V] to indicate that it follows a vowel stem. The verb ‘do’ is a so-called anomalous verb and as such is treated separately from regular verbs for most of its functions (see further §4.4.4.9 (c) below). But the infinitives of the anomalous verbs with vocalic stems, ‘go’, ‘do’ and ‘be’, seem to be no different in form, in any one text language, from the infinitives of other verbs with vocalic stems, whether these be original vowel stems, contracted or otherwise, or whether the stems have become vocalic from consonant loss. It seems reasonable therefore to include the infinitive endings of the anomalous verbs with the endings of other vowel stem verbs.33
If the infinitive is a complement following an infinitive marker (such as ‘to’, ‘forto’, ‘for’, ‘at’), -m is added to the basic label vi. In this way any surviving inflexion on the infinitive may be identified. Infinitive markers are given the label im. They are also marked for whether the verb following them begins with consonant (+C), <h> (+H) or vowel (+V):

$in{m}/pr_*ON
$3/qc<pr_yRE
{\}
$wi:se/npl<pr_wIS+E $/pln<pr_+E
$fast/vps13_FAST+Ed $/vps13_+Ed
$man/n_MAN
...
$/TN_yE
$sinful/aj_SYN+FULLE $-ful/xs-aj_+FULLE
$forto/im+C_FOR-TO
$cleanse/viK2-m_CLENS+EN $/viK2-m_+EN
$/P13OdXM_HIM
{.}
$/TN_yE
$righteous/aj_RIHTwISE
$forto/im+C_FOR-TO
$witian/viK2-m_wIT+IEND^E $/viK2-m_+IEND^E
$/P13GM_HIS
$righteousness/nOd_RIHTwIS+NESSE $-ness/xs-nOd_+NESSE
{.}
{\}
$&/cj_&
$nighlae:can/vi_NEH+LECH+EN $-lae:can/xs-vi_+LECH+EN $/vi_+EN
$to/pr+H_TO
$holiness/n<pr_HOLI+NESSE $-ness/xs-n<pr_+NESSE
{.}
$/TN_yE
$holy/aj_HOLIE
$man/n_MAN
$fast/vps13_FAST+Ed $/vps13_+Ed
$forto/im+H_FORTO
$high/vi-m_HEG+EN $/vi-m_+EN
$/P13GM_HI^S
{\}
$seat/nOd_SETE
$in{p}/pr_ON
$heaven/n<pr_HEUENE
{.}
$&/cj_&
$to/im+V_TO
$e:acan/vi-m_EK+EN $/vi-m_+EN
$/P13GM_HIS
$holiness/nOd_HOLI+NESSE $-ness/xs-nOd_+NESSE
{.}
$&/cj_&
$to/im+V_TO
$high/vi-m_EG+EN $/vi-m_+EN
$/P13GM_HI^S
$bliss/nOd_BLISSE


‘In three manners a man fasts ... the sinful to cleanse himself, the righteous to protect his righteousness and approach holiness. The holy man fasts to exalt his seat in heaven, to increase his holiness and to exalt his bliss’ (trhomBt.tag).

Contrast here the grammel for the verb $nighlae:can: the -m addition is only added to vi when the infinitive has its own infinitive marker not when it is the second of a pair of infinitives of which the first is marked. Sometimes the infinitive marker splits:

$all/av_*ALL
$for/pr_FOR^R
$/T<pr_yE
$love/n<pr_LUFE
$of/pr_OFF
$god/n<pr_GODD
{.}
$&/cj_&
$not/av_NOHHT
{.}
{\}
$for/im>=_*FOR^R
$earthly/ajOd_ERy+LIg $-ly/xs-ajOd_+LIg
$lof/nOd_LOFF
$to/im+C<=_TO
$winnan/vi-m_wINN+EN^N $/vi-m_+EN^N

‘... all for the love of God and not to gain earthly praise’ (ormt.tag).

Here *FOR^R and TO are treated like other coordinating constructions and are linked by >= and <= respectively.34

(b) Verbal noun
The basic label for the gerund or verbal noun is vn with the further addition, where relevant, of the extra verb labels K2, K2[1], F, Fier and Fir. The detached suffix is similarly marked:

{(& Sunt 3^A . Scilicet(}
$god/nG_GODD+ES $/Gn_+ES
$wrathful/aj_WREHT+FOL $-ful/xs-aj_+FOL
$werreien/vnF{rh}_WER\RA+YNGe $/vn{rh}_+YNGe
{.}
$man/nG_MAN+S $/Gn_+S
$rueful/aj_RU+FUL $-ful/xs-aj_+FUL
$outlaw/vnK2{rh}_UT+LAUH+YNG $out-/xp-v_UT+ $/vnK2{rh}_+YNG
{.}
$&/cj_AND
$/TG_YE
$devil/nG_DEU>E>L+S $/Gn_+S
$mightful/aj_MYHT+FUL $-ful/xs-aj_+FUL
$master/vnFier{rh}_MASTer+YNG $/vnFier{rh}_+YNG

And there are three. That is; the wrathful warring of God, the pitiful exile of man, and the powerful mastery of the devil’ (bodley26t.tag).

Apart from the leading v, identifying it as being derived from a verb, the grammel for a verbal noun is built up in other respects like the grammel for any noun. The genitive and plural endings are hived off as for other nouns, separately from the ‘-ung/-ing’ suffix. In the separate grammel of the final suffix the vn label is changed to n; this ensures that the genitive and plural endings may be grouped with other nominal suffixes:

$that/cj_y~
$/P13NM_HE
{\}
$teach/vsjps13_TEACHE
$/P22Od_OU
$against/pr_TO-zEINES
$fandian/vnpl<prK2_FOND+UNG+ES $/vnpl<prK2_+UNG+ES $/pln<pr_+ES

‘...that he may teach you against temptations’ (cleoarat.tag).

(c) Present participle
The basic label for present participle is vpsp with the further addition, where relevant, of the extra verb labels K2, K2[1], F, Fier and Fir. The detached suffix is similarly marked:

'_HEROUDES
$/TN_yE
$king/n_KING
$be/vpt13_WAS
$there/av{rh}_yERE
{\}
$herian/vpspK2[1]_*HER+IENDE $/vpspK2[1]_+IENDE
$/P23Od_HEOM
$as/cj_ASE
$/P23N_HUY
$god/npl_GOD+ES $/pln_+ES
$be/vpt23{rh}_WEREN

‘Herod the king was there, praising them as if they were gods’ (laud108at.tag).

If the referent of the participle is plural, the label -pl is added:

$/P23N_HI
$be/vpt23_WERE
$niman/vSpp-pl_I+NOM+E $ge-/xp-vpp_I+ $/vSpp-pl_+E
$handhaving/vpsp-pl_HOND-HABB+INGE $have/vpsp-pl-k_-HABB+INGE $/vpsp-pl_+INGE

‘They were taken red-handed’ (corp145selt.tag).

Participles are non-finite verb categories that are also adjectives. When they are used predicatively, as in the examples above, no further label identifying them as adjectives is added. This policy holds even when the referent of the participle is not the subject of the main clause and where the adjectival quality may therefore seem more obvious. However, when present participles are used as attributive adjectives this function is noted by the addition of the label -aj, the grammel being built up thereafter like any other adjective. Compare the grammels in the following two examples. In the first, the present participle is marked for plural but not as an adjective, because it is not used attributively. The second where both types are present, the first is marked as an attributive adjective, the second only as a plural present participle used predicatively:

$we:nan/vps12_*wEN+EST $/vps12[N]_+EST
$/P12N_yU
$/P23Od_HI
$bring/vi_BRING+E $/vi_+E
$so/av_SO
$lightly{w}/av{rh}_LIzT+LICHE $-ly/xs-av{rh}_+LICHE
{.}
{\}
$to/pr_*TO
$god/nG_GOD+ES $/Gn_+ES
$rich{k}/n<pr_RICHE
$all/av_AL
$sing/vpsp-pl{rh}_SING+INGE $/vpsp-pl{rh}_+INGE

‘Dost thou think to bring them so easily to the kingdom of God all singing?’(cotowlat.tag).

$&/cj_&
$all/ajpl_ALL+E $/plaj_+E
{\}
$/P23G_HEORE
$tear/npl_TER+ES $/pln_+ES
$be/vps23_BEOd
$burn/vpsp-ajpl_BERN+InDE $/vpsp-ajpl_+InDE
$gle:d/npl_GLED+EN $/pln_+EN
$gli:dan/vpsp-pl_GLID+EnDE $/vpsp-pl_+EnDE
$over/pr_OUER
{\}
$/P23G_HEORE
$own/ajpl<pr_AgEN+E $/plaj<pr_+E
$neb/npl<pr_NEBB+E $/pln<pr_+E

‘And all their tears are burning coals, gliding over their own faces’ (lamhomA1t.tag).

The detached suffixes of present participles are not further subdivided when in plural or genitive functions. See BERN+InDE above and the example below:

$dryhten/n-voc_*DRIHTEN
$/P12N_yU
{\}
$do/vps12_DE+ST $/vps12[V]_+ST
$/TOd_yE
$lof/nOd_LOF
$of{f}/pr_OF
$milkdrink/vpsp-ajplG_MILC-DRINK+ENDE $drink/vpsp-ajplG-k_-DRINK+ENDE
$/vpsp-ajplG_+ENDE
$child/nplG_CHILDRE
$mouth/n<pr_MUdE

‘Lord, thou brings praise out of the mouths of suckling children’ (lamhomA1t.tag).

When the participial adjective functions as a noun it is taken to be attributive with the following noun understood:

$/DisN_*yIS
$angel/n_ANGEL
$to/pr_TO
$/Dos<pr_yA
$quake/vpspK2-ajnpl<pr_Qu^AK+AnD $/vpspK2-ajnpl<pr_+AnD
$cy:Yan/vpt13{rh}_KID

‘This angel made known to the quaking [ones]...’ (edincmat.tag).

Note how the extra K2 label precedes the -aj label because refers to the verbal character of the word rather than its adjectival function.
Very occasionally, the present participle is used with a personal pronoun, in adverbial function, in the same way as a Latin ablative absolute. In the example form Ayenbite of Inwyt below, it actually translates Latin te loquente. This is not normal Middle English syntax and the extra label -aa-, for ablative absolute, has been assigned to this special usage. The label -av added to grammel of both pronoun and participle indicates adverbial function:

$right/aj_RYzT
$/P13NI_HIT
$be/vps13_IS
{,}
$that/cj_yET
$/P12-av_yE
$speak/vpsp-abs-av_SPEK+INDE $/vpsp-aa-av_+INDE
{*translated absolute*}
{,}
$well{w}/av_WEL
{\}
$more/av_MORE
$/P21N_WE
$be/vsjps21_BY
$still/ajpl_STILL+E $/plaj_+E

‘It is right that, with you speaking, we should be the more quiet’ (ayenbitet.tag).

$that/cj_*yAT
$/P22N_GIE
$/P13OdI_IT
$do/vpt22_DIDE
$/P22G-av_GIUR
$unwitan/vpsp-abs-av{rh}_UN+WIT+AnD $un-/xp-v_UN+ $/vpsp-aa-av{rh}_+AnD
{*reads like a translated ablative absolute*}

‘...that ye did it unknowingly [lit: with your unknowing]’ (edincmct.tag).

(d) Weak past participle
The basic label for the past participle of weak verbs is vpp. This is also the default grammel for any past participle that is not formally strong, e.g. weak forms of the past participle found in verbs that are normally strong in early Middle English and also for irregular past tenses such as ‘did’ and ‘ēode’ (see further §54.4.9.3 below). Weak past participles are formally similar to weak past tense indicatives and their endings are hived off in the same way. As with vps12, vps13 and vpt (see §4.4.4.2 above), detached suffixes are marked with specifiers for stem-final consonants where these have a potential effect on their form, whether causing syncopation or devoicing or both. The participles and their detached suffixes may carry the further addition, where relevant, of the extra verb labels K2, K2[1], F, Fier and Fir.

$/P23N_HE
$have/vps23-apn_HAU+EN $/vps23-apn_+EN
$hear/vpp_HER+D $/vpp[R]_+D
$tell/vpp_TOL+D $/vpp[L]_+D
$of/pr_OF
$/Dis<pr_dIS
$mermaid/n<pr_MERE

‘They have heard tell [told] of this mermaid’ (bestiaryt.tag).

When the past participle is used with the copula, if its is plural, the label -pl is added:

$there/av_yER
$be/vps23_BYEy
...
$town/npl_TOUN+ES $/pln_+ES
$forburn/vpp-pl_UOR+BERN+D $for-/xp-v_UOR+ $/vpp[N]-pl_+D
{.}
$abbey/npl_ABBEY+ES $/pln_+ES
{.}
$priory/npl_PRIORI+ES $/pln_+ES
{.}
{\}
$barn/npl_BERN+ES $/pln_+ES
$destroy/vpp-plF_DESTRU+D $/vpp[V]-plF_+D
{.}
$&/cj_AND
$man/npl_MEN
{,}
$&/cj_AND
$woman/npl_WYFMEN
{,}
$&/cj_AND
$child/npl_CHILDREN
$disinherit/vpp-plF_DESERIT+ED $/vpp-plF_+ED
{,}
$&/cj_AND
$exile/vpp-plFier_Y\-EXIL+ED $ge-/xp-vpp_Y\- $/vpp[L]-plFier_+ED

‘There are ... towns burnt down, barns destroyed, and men, women and children disinherited and exiled’ (ayenbitet.tag).

A special label is used for a phenomenon that occurs twice in the manuscript of Ancrene Wisse; this is where superscript <e> used as an abbreviation for <re> is part of the past participle suffix. In these cases the vpp label is given the addition [r] followed by [R] to indicate that the ^E of the suffix includes the notional <r> of the stem the expansion of the abbreviation for which appears preceding the +:35

$me/int_*ME
$which/aj_HwUCH
$unsae:lig/aj_UN+SELI $un-/xp-aj_UN+
$ghost/n_GAST
$have/vps13_HAU+Ed $/vps13_+Ed
$so/av_SwA
$bimalscren/vpp_BI+MALSCr+^ET $be-/xp-v_BI+ $/vpp[r][R]_+^ET
{=^E stands for <re> so is part of the stem as well as the ending=}
{\}
$/P22Od_Ow

‘Alas what unhappy spirit has so bewitched you?’

The weak past participle may also be assigned the label aj, like the present participle, when it is used as an attributive adjective:

$so/av_SwA
$/P12N_dU
$do/vpt12_DEDEST
$/TOi_dE
$forguilt/vpp-ajOi_FOR+GILTE $for-/xp-v_FOR+
'_MARIE-MAGDALENE

‘So thou didst to the guilty Mary Magdalene’ (vvbt.tag)

As with the present participle, the past participle used as a noun is counted as an attributive adjective:

$go/v-imp_*GA
$away/av_AwEI
$from/pr_FRAM
$/P11<pr_ME
$/P12-voc_dU
$waergan/vpp-ajn-voc_gE+wEREg+EDE $ge-/xp-vpp-ajn_gE+ $/vpp[J]-ajn-voc_+EDE

‘Go away from me thou cursed [one]’ (vvat.tag).

4.4.4.9 Strong, preterite present and anomalous verbs
In some categories, the suffixes of strong and weak verbs are identical, and no differentiation is therefore made between their grammels for these categories, viz: $/vps1, $/vps2, $/vi, $/vpsp, $/vn, $/vsjps and $vsjpt (except for 1st and 3rd singular for which see §4.4.4.6 above). The grammels for strong verbs in these categories will by definition of course never carry the extra labels K2, K2[1], F, Fir and Fier. For some categories, however, strong verbs behave differently from weak verbs and for these categories their grammels carry the extra label S immediately following v.

4.4.4.9.1 Strong verbs — specially marked categories
(a) Past tense
What differentiates strong from weak verbs is ablaut, which affects stem vocalism of both singular and plural. The singular forms have no detachable suffix; the base form is labelled vSpt so that strong past tense forms may be identified and retrieved separately from those of weak verbs:

$/TN_*yE
$queen/n_QUENE
$of/pr_OF
;_*SABA
$come/vSpt13_COM
$thereforth/av-k_yer-UORy $forth/av-k_-UORy
{.}
$&/cj_&
$anonso/cj_ANON-SO
$/P13NF_HE[O]
$/P13OdI_IT
$see/vSpt13{rh}_I+SEY $ge-/xp-vpt13_I+
{\}
$honour/vpt13F_*HONOUR+EDE $/vpt13[R]F_+EDE
$/P13OdI_IT
$fair/av_FAIRE
$&/cj_&
$sit/vSpt13_SAT
$aknee/av_AKNE[O]

‘The queen of Sheba came forth there, and as soon as she saw it she honoured it duly and went down on her knees’ (corp145selt.tag).

In Old English, strong past tense plurals historically had the suffix -on. This survives in early Middle English as vowel (normally reduced to <e>) with or without final <n>. These endings are therefore hived off, and in instances where the ending is completely lost, the ‘form’ of the hived off ending is given as 0 for comparison. The only stem finals that potentially could cause syncopation are vowel and <r>. These are therefore marked in the relevant grammels, though in contrast to their effect on the forms of the endings of the strong past participle, where syncopation occurs quite frequently,36 it appears that they almost never cause syncopation in the strong past tense:

$/TplN_*yE
$ploughman/npl_PLOU+MEN $man/npl-k_+MEN
$well{w}/av_WEL
$/DatpnOd_yAT
$see/vSpt23{rh}_*I-SEY+E $ge-/xp-vpt23_*I- $/vSpt23[V]{rh}_+E
{\}
$&/cj_*AND
$blow/vSpt23_BLEW+EN $/vSpt23_+EN
$out/av_HOUT
$horn/nOd_HORN
$in{p}/pr_IN
$/T<pr_yE
$le:ah/n<pr{rh}_LEYE

‘The ploughmen saw that clearly and gave out a horn call in the meadow’ (digby86mapt.tag).

(b) Past participle
The basic label for strong past participle is vSpp. The historical ending (OE -en) may survive intact in early Middle English, or be reduced to <e> or to zero. The endings are hived off in the same way as those of the past tense plural with vowel-stems and <r>-stems marked:

$bu:tan/cj_BUTEN
$/P13NI_HIT
{\}
$be/vsjps13_BEO
$/D-cpv_yE
$better/av_BET+ERE $-er/xs-cpv_+ERE
$hide/vpp_I+HUD $ge-/xp-vpp_I+
$&/cj_&
$helan/vSpp_I+HOL+EN $ge-/xp-vpp_I+ $/vSpp_+EN
{.'}
$/P13NI_HIT
$be/vps13_IS
$forlose/vSpp_FOR+LOR+EN $for-/xp-v_FOR+ $/vSpp[R]_+EN
$soon/av_SONE

‘Unless it be the better hidden and concealed, it is soon lost’ (caiusart.tag).

$&/cj_*AND
$when/cj_QUEN
$/P12N_yOU
$have/vps12_HAU+IS $/vps12_+IS
$/P12OdX_yE-SELUEN $-self/xs-Od_-SELUEN
$slay/vSpp{rh}_SLA+N $/vSpp[V]{rh}_+N
{\}
{~f23vb~}
$to/pr+H_*TIL
$heaven/n<pr_HEUIN
$shall/vps11_SALLE
$/P11N_*I
$gera/vi_GER $/vi_0
{[thee[}
$be/vi_BE $/vi[V]_0
$take/vSpp{rh}_TA+NE $/vSpp[V]{rh}_+NE

‘And when thou hast killed thyself I shall cause thee to be taken to heaven’ (edincmbt.tag).

Like the present participle and the weak past participle, the strong past participle is marked with aj when used attributively:

$child/npl_*CHILDRIN
$unborn/vSpp-ajpl_ON-BOR+N $un-/xp-v_ON- $/vSpp[R]-ajpl_+N
$/RTApl_yAT
$naught/nOd_NAUT
$/neg-v(<n)_NE
$see/vps23{rh}_SE+y $/vps23[V]{rh}_+y

‘Unborn children who see nothing...’ (adde6bt.tag).

4.4.4.9.2 Preterite present verbs
The preterite present verbs that survive into early Middle English are listed in Campbell (1959: §767) as OE: wāt ‘know’, āh ‘possess’, dēag ‘avail’, ann ‘grant’, cann ‘can, know’, þearf, ‘need’, dearr ‘dare’, sceal ‘shall’, mōt ‘must’ and mæg ‘may’. In the LAEME corpus these are identified by the following lexels: $witan, $a:gan, $dugan, $unnan, $can, $Yurfan, $dare, $shall, $mo:t, and $may.37 These verbs are characterised by the forms of their present tense having a strong past form. The forms of these verbs are therefore for the most part not strictly comparable with either weak or strong verbs. They are marked as usual for tense, number and person, and for any non-finite categories that occur. Most potentially detachable endings, however, are left undetached: once endings are hived off from the parent word, they become anonymous when listed as suffixes in the text dictionaries. Non-comparable categories should therefore not be mixed with those that are strictly comparable.

$not/av>v_NAwT
$only/av_ANE
$for/pr_FOR
$/P12<prX_yE-SEOLF $-self/xs-<pr_-SEOLF
{.'}
$as/av_ASE
$Yurfan/vps13_yE\RF
$god/nG_GOD+ES $/Gn_+ES
$spouse/n_SPUSE
{.'}
$ac/cj_AH
$shall/vps12_SCHALT
$for/pr_FOR
$many/ajpl<pr_MONI+E $/plaj<pr_+E
$other/pnpl<pr_OyR+E $/plpn<pr_+E

‘...not only for thyself, as God’s spouse needs to do, but [thou] shalt [keep vigil] for many others’ (bod34t.tag).

$&/cj_ANT
$/P21N_wE
$/neg-v_NE
{\}
$may/vps21_MAHEN
$/DatpnOd_yET
$do/vi_DO+N $/vi[V]_+N
{.’}
$then/av_yENNE
$darian/vps21-bpn_DARI+E $/vps21[V]-bpn_+E
$/P21N_wE
{.}
$&/cj_&
$/neg-v(>av)_NE
$dare/vps21_DURREN
$never/av_NEUE*R
$come/vi_CUM+EN $/vi_+EN
{\}
$before{p}/pr_BI-UOREN
$/P13<prM_HIM

‘...and we may not do that. Then we hide and never dare come before him’ (royalkgbt.tag).

$/neg-v(>av)_*NE
$Yurfan/vpt13_yURHTE
{*sj context*}
$/P12Oi_yE
$never/av_NEUer
$rue/vi{rh}_REW+E $/vi{rh}_+E
$may/vpt12+P_MYHTES+TU $/P12N-as_+TU
$do/vi_DO $/vi[V]_0
$/P12OdX_yE
$in{m}/pr_IN
$/P13GM_HIS
$hyldu/n<pr{rh}_YLDE

‘If thou might put thyself in his protection, you would never need to regret it’ (jes29t.tag).

For the preterite-present verbs, the exceptions to this rule are forms, where they occur, of the imperative plural, the infinitive, the verbal noun and the present participle, in all of which the endings seem to be the same as they are for any other kind of verb, and may therefore be sensibly directly compared with them.

$/neg-v(>av)_*NE
$may/vps13_MAI
$never/av_NEUERE
$ma:nswara/n_MON-SWARE $swara/n-k_-SWARE
{\}
$manship/nOd_MON-SCIPE $-ship/xs-nOd_-SCIPE
$long/av_LONGE
$a:gan/vi_Az+En $/vi_+En
{MS AzEnT with T subpuncted}

‘A perjuror may never have honour long’ (layamonAbt.tag).

$run/v-imp22_IERN+Ed $/v-imp22_+Ed
$to/pr_TO
$/T<pr_dE
$tree/n<pr_TREwE
$/RTI_dE
$stand/vps13-ct_STANT
$amidward/pr_AMIDEwARDE
$paradise/n<pr_ParADISE
$witan/vpsp_wIT+EnDE $/vpsp_+EnDE
$both/av>=_BOdE
$good/nOd_GOD
$&/cj<=_&
$evil/nOd_EUEL

‘Run to the tree that stands in the middle of Paradise, knowing both good and evil’ (vvat.tag).

In addition, the verb $unnan has regularly-formed strong past participle endings (and occasionally also weak past participle endings) which are hived off in the usual way.

$/RTIN-Od_yAT
$/P21G_VRE
$foregengel/npl_FORD+GENGL+ES $fore-/xp-n_FORD+ $/pln_+ES
$unnan/vpt23_VTHEN
$&/cj_AND
$/RTIN-Od_yAT
$/P21N_WE
$unnan/vSpp_VNN+EN $/vSpp_+EN
$have/vps21_HABB+ETH $/vps21_+ETH
$into/pr_INTO
$/T<pr_yAT
$holy/aj<pr_HOLI
$minster/n<pr_MINSTER

‘...that our predecessors granted and that we have granted to the holy minster’ (buryFft.tag).

Contrast undivided past tenseVTHEN and past participle VNN+EN with the ending hived off.

4.4.4.9.3 Anomalous verbs
The anomalous verbs, as their name implies, are a more disparate and irregular group: $will (along with $willnot),38 $do, $go,39 and $be (along with $benot).40 To this may be added the modal auxiliary mun, mon ‘must’, derived from Scandinavian, and given the lexel $munu. The forms of $will, $willnot and (rare in early Middle English) $munu, like the preterite-present verbs used as modal auxiliaries, do not have their endings hived off but are kept as single units comparable only with the variant forms sharing the same tag.

$/P23N_HA
{\}
$will/vpt23_wALDEN
$catch/vi_KECCH+EN $/vi_+EN
$of{f}/pr_OF
$/P12<pr_yE
$all/ajnOd_AL
$/RTIOd_yAT
$/P12N_yU
$have/vpt12_HEAU+EDEST $/vpt12_+EDEST

‘They would catch from thee all that thou had’ (caiusart.tag)

$with{w}/pr_*WIT
$hard/ajpl<pr_HARD $/plaj<pr_0
$dynt/npl<pr_DINT+IS $/pln<pr_+IS
$munu/vps23-bpn_MON
$/P23N_yAI
$cy:Yan/vi{rh}_KIy+E $/vi{rh}_+E

‘With hard blows they make known...’ (edincmat.tag).


$do and $go are treated as vocalic stem strong verbs for present tense, infinitive, verbal noun and present and past participle, where their forms are comparable to other vocalic stem strong verbs.

$/P13NM_*HE
$will/vps13_wILE
$go/vi_GO+N $/vi[V]_+N
$in{p}/pr_A+
$right/aj<pr_+RIzTE
$way/n<pr{rh}_wEIE

‘He will go in the right way’ (cotowlat.tag).

$say/v-imp22_*SEGG+Ey $/v-imp22_+Ey
$/P11Oi_ME
$who/pn_HWO
$have/vps13_HAU+Ey $/vps13_+Ey
$/DispnOd_yIS
$do/vSpp{rh}_I-DO $ge-/xp-vpp_I- $/vSpp[V]{rh}_0

‘Tell me who has done this’ (jes29t.tag).


Their past tense forms, however (those of $go under the separate lexel $e:ode) are not considered to be always comparable to those of other verbs and are therefore preserved unsegmented.

$/P23N_*HII
$fe:ran/vpt23_VER+DEN $/vpt23[R]_+DEN
$ever/av_EF\RE
$forthright/av{rh}_FORy-RIHT
$before{p}/pr_BI-FORE
;_*AFFRIKE
$/P23N_HII
{\}
$e:ode/vpt23_EODE
$forth/av{rh}_FORy
$&/cj_AND
$ever/av_EUERE
$/P23N_HII
$draw/vSpt23_DROW+E $/vSpt23_+E
{\}
$west/av_WEST
$&/cj_AND
$north/av{rh}_NORy

‘They journeyed ever straight ahead; along the coast of Africa they went forth, and always they made west and north’ (layamonBOt.tag).


$be is treated as a vocalic stem strong verb for past participle and infinitive only:

$when/cj_*dANNE
$/P21N_wE
$have/vps21-apn_HAU+EN $/vps21-apn_+EN
...
$be/vSpp_BE+N $/vSpp[V]_+N
$hu:slian/vpp-plK2_HU\SEL+ED $/vpp[L]-plK2_+ED
{.}
$/P21N_wE
$be/vps21-apn_BEN
$high/ajpl_HEG+E $/plaj_+E

‘When we have been given the eucharist, we are high...’ (trhomBt.tag).

Note the past participle with hived off ending and the plural indicative left undivided.

$fast/vps13_FAST+Ed $/vps13_+Ed
$/TN_yE
$holy/aj_HOLIE
$man/n_MAN
$for/im_FOR
$be/vi-m_BE+N $/vi[V]-m_+N
$high/aj_HEGH
{\}
$in{p}/pr_ON
$heaven/n<pr_HEUENE

‘The holy man fasts to be high in heaven’ (trhomBt.tag).

There is one special grammel for $be. The verb is sometimes used to express futurity. In these cases it is always a form of ‘be’ that is used, rather than a form of ‘art’, ‘is’ or ‘are’, and the grammel is v-fut followed by the usual number and person labels:

$then/av_yENNE
$be/v-fut13_BId
$/TN_yA
$evil/aj_UFELE
$sorriness/n_SARI+NESSE $-ness/xs-n_+NESSE
$withal/av_MID-ALLE
{\}
$overcome/vi_OUER+CUM+AN $over-/xp-v_OUER+ $/vi_+AN
$mid{w}/pr_MID
$/P21G_URE
$good/aj<pr_GODE
$Yyld/n<pr_I+yULDE $ge-/xp-n_I+
{.}
$/TN_*yE
$6/qoaj_SIXTE
{\}
$might/n_MIHTE
$be/vps13_IS
{.}
{[instantia boni operis[}
$/Datpn_y~
$be/vps13_IS
$a:nrae:dness/n_ANRED+NESSE $-ness/xs-n_+NESSE
$good/ajplG_GOD+ES $/plajG_+ES
$work/nplG_wER\K+ES $/plnG_+ES

‘Then shall the evil sorrow withal be overcome through our good endurance. The sixth virtue is [instantia boni operis] that is diligence in good works’ (lamhomA1t.tag).

Note the contrast between future BId and existential IS, though this distinction is not absolute in this text language, BId also being used where no futurity is implied.


4.4.5 Clausal Negation41

4.4.5.1 The main types of clausal negation
In early Middle English the formation of negative clauses falls for the most part into one of three main types: (1) the negative adverb ‘ne’ immediately preceding the verb, continuing Old English usage; (2) the verb preceded by ‘ne’ and followed by another negative adverb ‘not’; (3) ‘not’ alone following the verb. These three types represent the three stages in the historical development of negative clauses characterised as ‘Jespersen’s Cycle’ or the Negative Cycle (Jespersen 1962 [1917]: 9–14; 1924: 335–336). The tags have been devised to take account of these three main constructions.44 In type (1) the preceding verbal negator is not given a lexel, but the grammel $/neg-v is assigned to the simple negative marker which is usually spelled NE, though NA, NI and NO are also found:

$/P23N_*HI
$/neg-v_NE
$reccan/vps23_RECCH+Ey $/vps23_+Ey
$how/av_HU
$/P23N_HI
$live/vsjps23{rh}_LIBBE

‘They do not care how they live’ (cotowlat.tag)

The following example shows how the ‘ne’ may be cliticised to the following finite verb. So-called ‘negative contraction’ most often occurs in forms of the verbs ‘āgan’ (= ‘own, ought’), ‘be’, ‘have’, ‘will’, ‘witan’ (= ‘know’) that begin with a vowel, or with <h> or <w> followed by a vowel. In these cases the lexel of the negative verb ends in ‘not’ (e.g. $havenot, $willnot, $witannot) and the cliticised ‘ne’ is characterised as a verbal prefix and given the tag $n-/xp-neg.

$yet/av_*zET
$benot/vps13_N+IS $n-/xp-neg_N+
$/DisN_yOS
$speech/n_SPECHE
$bring/vpp_I+BROz+T $ge-/xp-vpp_I+ $/vpp[H]_+T
$to/pr_TO
$doom/n<pr{rh}_DOME

‘This debate is not brought to judgement yet’ (cotowlat.tag)

Type (2), ‘ne + not’ is known as split or braced negation. The tagging is designed to isolate the two elements of the verbal negation, and then to combine them as a coordinating construction in the text dictionaries and the maps. Braced negation is therefore tagged in a similar way to other co-ordinating constructions (see further §4.4.6.3 below, and cf. the treatment of discontinuous compound relatives described in §4.4.2.3 above).

$/P12N_yOU
$/neg-v>=_NE
$shall/vps12_SSELT
$not/neg-v<=_NAzT
$consent/viFir_conSENT+I $/viFir_+I
{,}
$to/im+C_TO
{\}
$do/vi-m_DO $/vi[V]-m_0
$sin/nOd_ZEnNE
{,}
$mid{w}/pr_MID
$/P12G<pr+C_yINE
$body/n<pr_BODYE

‘Thou shalt not consent to do sin with thy body’ (ayenbitet.tag)

The first element is similar to type (1), and it is given the same tag but has >= added to its grammel: $/neg-v>=. This indicates that it refers to the negative adverb ‘ne’, but the directional arrow shows that it points towards another negator and ‘=’ that it coordinates with it. The second element is given the tag $not/neg-v<=. The lexel indicates that the negative word is ‘not’, while the grammel and following <= show that it negates the finite verb and coordinates with the preceding ‘ne’.43 Braced negation may also appear with negative contraction:44

$Ya:/cj_*yO
$/DisN_yIS
$emperor/n_EMPerOR
$see/vSpt13_I+SEI $ge-/xp-vpt13_I+
{.}
$that/cj_y^T
$/P13NM_HE
$willnot/vpt13_N+OLDE $n-/xp-neg>=_N+
$not/neg-v<=_NOzT
$turn/vi_TurN+E $/vi_+E
$/P13GM_IS
$thought/nOd{rh}_yOzT

‘When this emperor saw that he would not turn his thought ’ (corp145selt.tag)

The tag $not/neg-v<= always indicates the second element of braced negation. It is assigned therefore, in this context, not only to all the numerous variant spellings in the LAEME corpus of ‘not’ (< OE )(n)wiht) but also to the forms NA and NO (< OE )) when used in this construction:

$Ya:/cj_yA
$/P13NM_HE
$/neg-v>=_NA
$spare/vpt13K2_SPar+EDE $/vpt13[R]K2_+EDE
$not/neg-v<=_NA
'_IHesU
$christ/nOd-t_Cr^IST
$/P13GM_HIS
$own/ajOd_AgENE
$son/nOd_SUNE

‘When he spared not Jesus Christ, his own son’ (lamhomA1t.tag)

Type (3), ‘not’ alone after the verb, is a later construction that begins to displace the other two in late Middle English. But it is already present in some of the LAEME texts and is assigned the tag $not/neg-v<. This is similar to the tag for the second element of braced negation, and indicates that a form of the word ‘not’ (or ‘na/no’) negates the clause. The directional arrow indicates that the ‘not’ follows the finite verb, not necessarily contiguously, while the absence of ‘=’ indicates that there is no coordination with another verbal negator:

$listen/v-imp22_*LESTN+IT $/v-imp22_+IT
$now/av_NOU
$&/cj_AND
$be/vps22_BETz
$not/neg-v<_NAzT
$deaf/ajpl{rh}_DEU+E $/plaj{rh}_+E

‘Listen, now, and be not deaf’ (adde6bt.tag)

4.4.5.2 Minor types of verbal negation
In the LAEME corpus there are three further types of verbal negation that seem to be synonymous with the three main types: that is, they appear merely to negate the clause without any extra negative emphasis implied.

(i) Very rarely, ‘not’ alone precedes the verb and this is indicated by placing the directional arrow between ‘neg’ and ‘v’ in the grammel as in the example below:

$/P13GM_*HIS
$name/nOd_NAME
$not/neg>v_NAzT
$swear/v-impK2[1]_SUERIE
$in{m}/pr_AN
$idleship/n<pr{rh}_YDEL+SCHUPE $-ship/xs-n<pr{rh}_+SCHUPE

‘Do not take his name in vain’ (emmanuel27t.tag)

Also very uncommon is (ii) ‘ne + ne’ plus finite verb45 and rather more frequently found is (iii) ‘not + ne’ plus finite verb (Laing 2002: 309–313).46

The three examples of type (ii) given below show that this construction may be found with or without negative contraction, and that the two preverbal negators are not always adjacent in the clause:

$if/cj_*IF
$woman/n_WYMMOn
$think/vps13_yENCH+Ey $/vps13_+Ey
$love/viK2_LUU+YE $/viK2_+YE
$derne/av{rh}_DERNE
{.}
{\}
$/neg>neg>=>v_*NE
$/neg-v<neg<=_NE
$may/vps11_MAY
$/P11N_IC
$/P11GOdpl+C_MYNE
$song/nplOd_SONG+ES $/plnOd_+ES
$wernan/vi{rh}_WERN+E $/vi{rh}_+E

‘If a woman thinks to love secretly, I may not withhold my songs’ (jes29t.tag)

Here the tags reflect the construction closely. The first ‘ne’ is labelled $/neg>neg>=>v, indicating that it is a negator preceding another negator with which it coordinates and both precede the verb. The second is labelled $/neg-v<neg<=, indicating a verbal negator preceded by another negator with which it coordinates.47

$though/cj_*yEIH
$/P21N_wE
$/P23Od_HES
$/neg>neg>=>v_NE
$seenot/vSpt21_N+ISEI+EN $n-/xp-neg<neg<=_N+ $/vSpt21[V]_+EN
$/P23N_HIE
$be/vpt23_wAREN
$/P21G_URE
$gefe:ra/npl{rh}_I+FER+EN $ge-/xp-n_I+ $/pln{rh}_+EN

‘Though we did not see them, they were our companions’ (trinpmt.tag)

Here the tag for the cliticised negator is treated as a verbal prefix, like that for the simple preverbal ‘ne’, but otherwise is given the same grammel as that for the uncliticised second element.48 The third example shows how the two parts of the negation may be separated by an intervening pronoun:

$&/cj_&
$/RTApl_dO
$/neg>neg>=>v_NE
$/P13OiI_HIRE
$/neg-v<neg<=_NE
$will/vps23_WYLE
$listen/vi_I+LEST+En $ge-/xp-vi_I+ $/vi_+En
$/P23N_HIE
$become/vps23-apn_BE+CUM+Ed $be-/xp-v_BE+ $/vps23-apn_+Ed
$to/pr+V_TO
$unmight/n<pr_UN+MIHTE $un-/xp-n_UN+

‘And those who will not listen to it, they become powerless’ (vvat.tag)

The four examples below of the commoner type (iii), preverbal ‘not + ne’, indicate that (like ‘ne + ne’) it may be found with or without negative contraction, and that ‘ne’ and ‘not’ need not be adjacent.

$/P13NM_HE
$beseech/vpt13_BE+SOH+TE $be-/xp-v_BE+ $/vpt13[H]_+TE
$at/pr_AT
$god/n<pr_GODE
$that/cj_yAT
$not/neg>v>=_NAHT
$/neg>v<=_NE
$shall/vpt13_SCOLDE
$rain/viK2_REIN+IN $/viK2_+IN
$for/im>=_FOR
$/TOd_dE
$folk/nOd_FOLKE
$to/im+C<=_TO
$chasten/viFier-m_KAST+IN $/viFier-m_+IN

‘He besought God that it should not rain, to chastise the people’ (vvbt.tag)

Here the first element of the double negation is tagged $not/neg>v>=. This follows the pattern of the simple preverbal ‘not’ in type (i) above ($not/neg>v), but the added >= indicates the presence of a following negator with which it coordinates. The second element is formally and functionally identical to the second element of type (ii) above — ‘ne + ne’ ($/neg-v<neg<=); but for operational reasons it requires to be given a different tag to indicate that it is part of a different coordinating structure. The tag $/neg>v<= indicates that it is a verbal negator appearing before the verb but coordinating with a preceding ‘not’.49
In the example above it might be argued that NAHT could be read as the emphatic adverb ‘not at all’. But there are similar examples where it is clear that ‘not + ne’ in combination is used to translate the Latin sentential negator non where no special emphasis is implied:

$forthat/cj_FOR-dAN
$/P12N_yU
$shall/vps12_SCALT
$be/vi_BIE+N $/vi[V]_+N
$dead/aj_DEAD
$&/cj_&
$not/neg>v>=_NAHT
$/neg>v<=_NE
$shall/vps12_SCALT
$/P12N-ad_TU
$live/vi_LIBB+EN $/vi_+EN

‘For thou shalt be dead and thou shalt not live’ (vvbt.tag)

This translates the Latin: quia morieris tu & non uiues. Compare also the example below, illustrating the ‘not + ne’ construction with negative contraction,50 which translates the Latin: & ipse nescuit:

$&/cj_&
$/P13NM_HE
$/P13OdI_HIT
$not/neg>v>=_NAwT
$witannot/vpt13_N+USTE $n-/xp-neg<=_N+

‘And he did not know it’ (tituslang2t.tag)

In these examples it is clear that the ‘not + ne’ construction serves as simple verbal negation and the ‘not’ is therefore tagged here, and in other examples where no special emphasis is implied, not as the emphatic adverb but as the first element of a form of double negation.51 The next example illustrates how the two negators may be divided by a personal pronoun:

$how/av_*HU
$may/vps11_MAI
$/P11N+V_IC
$in{p}/pr_On
$/T<pr_dANE
$world/n<pr_wORLD
$wunian/viK2_wUN+IgEn $/viK2_+IgEn
$&/cj_&
$not/neg>v>=_NAHT
$/P13OdI_HES
$/neg>v<=_NE
$love/viK2_LUU+IgEn $/viK2_+IgEn

‘How may I live in the world and not love it?’ (vvat.tag)

Adverbs may also intervene between the two parts of the negation and when this happens it seems that the ‘ne’ element serves to negate the clause and the preceding ‘not’ element has a double function — reinforcing the negation and flagging its scope:

$many/ajnpl_*MANIg+E $/plajn_+E
$say/vps23_SEGG+Ed $/vps23_+Ed
$/P13NM_HE
$say/vpt13_SA+DE $/vpt13[V]_+DE
$/RTApl_dE
$not/neg>v>=_NOHT
$well{w}/av_wEL
$/neg>v<=_NE
$believe/vps23_BE+LIEU+Ed $be-/xp-v_BE+ $/vps23_+Ed

‘“Many say”, he said, “who do not believe well”’ (or ‘ “who believe not well”’) (vvat.tag)

Middle English, like modern English, uses sentential negation even when the scope of the negation is not the main clause. Unlike modern English, early Middle English can utilise the separating capacity of double negation to indicate that the scope of the negation is phrasal. Here, the NE part of the negation negates the clause while the NOHT part flags the scope of the negation. These people do believe but they don’t do it well.52
There is a single example in the LAEME corpus53 of a construction that appears to combine preverbal ‘ne + ne’ (type (ii) above) with type (2) in §4.4.5.1 above — ‘ne’ preceding and ‘not’ following the verb. The tagging system is very flexible and there are a number of ways that this unusual structure could be signalled.54 The solution offered below shows how the tags of the two outer elements indicate the presence of, and point to, each other; but they also indicate the presence of, and point to, the middle element. The tag of the middle element in turn indicates the presence of both preceding and following coordinating negators:

$man/indef_ME
$/neg>neg>v>=_NE
$/P13XM_HIm
$/neg-v<neg<=>neg>=_NE
$aslacken/vsjps13K2_A\+SLAKY $a-/xp-v_A\+
$not/neg-v<neg<=(>nor)_NAzT
$too/av_TO
$much/av_MOCHE
$/TOd_yANE
$bridle/nOd_BRIDEL
$to/im+C_TO
$run/vi-m_YERN+E $/vi-m_+E
$to/pr_TO
$lust/npl<pr_LOST+ES $/pln<pr_+ES
$of/pr_OF
$/T<pr_yE
$flesh/n<pr_ULESSE
{,}
$nor/cj_NE
$to/pr_TO
$/T<pr_yE
$covetise/n<pr_CO\UAYTISE
$of/pr_OF
$/Dis<pr_yISE
$world/n<pr_WORDLE

‘That a man should not, with regard to himself, slacken too much the bridle so as to yearn towards lusts of the flesh nor towards the covetousness of this world’ (ayenbitet.tag)

In the example above, note also the tag to the third element, NAzT, of the sentential negation: $not/neg-v<neg<=(>nor). After <=, which indicates that the element coordinates with the preceding elements, there appears (>nor) as an addition to the tag. The bracketed addition points forward to the negative conjunction ‘nor’ that links two phrases dependent on the main negative construction. This leads us to discussion of how multiple negation is signalled in the LAEME tagging system.

4.4.5.3 Multiple negation
The three main types of verbal negation, and the minor types detailed in §4.4.5.2 above, are marked by their designated tags, and so differentiated in the text dictionaries. The types may then easily be identified and isolated for comparison and mapping. For the same purpose, the various kinds of braced negation, once identified, then have their tags and forms combined as coordinating elements. But these types of sentential negation do not tell the whole story. We also wish to indicate in the tagging the presence and extent of multiple negation from so-called ‘negative concord’. This is the syntactic rule by which certain words in a negated clause take on a negative form (in concord with the negator of the clause) and do not cancel each other out. These further negative words may be adverbs (‘never’, ‘nowhere’), adjectives (‘no’), nouns (‘nothing’, ‘no-man’), pronouns (‘none’, ‘naught’ (see further §4.4.5.5 below)), the coordinating sequence ‘neither..nor’ or the conjunction ‘nor’ (see further §4.4.5.6 below).
Sometimes a large number of negative words may accumulate in a negated clause; but (with the exception of the conjunction ne ‘nor’) only rarely are further negative words found in braced negation containing the element ‘not’ (‘ne + not’ or ‘not + ne’ types). The operation of this general syntactic rule was first stated by George Jack (1978: 299), and in
Laing (2002: 299–301) it is designated ‘Jack’s Law’.
The phenomenon of multiple negation, and its modification in terms of Jack’s Law, is clearly of great interest to the history of negation in English. The tags on verbal negators have therefore been expanded so that they can signal the presence of further negative words within the clause governed by the negated finite verb, and also give indication of their number and type. However, as the basic types of verbal negation described in §4.4.5.1 and §4.4.5.2 above need to remain comparable in the text dictionaries and the maps, whether or not there are also other negative words present, these additions to the tags are put in parentheses. The material within the parentheses will always be visible in the tagged texts themselves, where it can be observed in context, but it can be switched ‘on’ or ‘off’ when it comes to making text dictionaries or creating maps. When the material in parentheses is switched off, the verbal negation tag will be identical to whichever type it matches (e.g. preverbal ‘ne’, ‘not + ne’, ‘ne + ne’ etc), and can be amalgamated with others of the type in a particular text dictionary and/or the forms associated with it mapped at the locations of the relevant texts. But if the material in parentheses is switched on, it is thereby possible to get a general idea of the extent of the use of multiple negation in that particular text. The examples below illustrate the complete tags with all material in parentheses included.

$for/cj_FOR
$there/av_yER
$fire/n_FUR
$/neg-v(>aj)_NE
$give/vps13_zEU+Ed $/vps13_+Ed
$no/ajOd+C_NA
{\}
$light/nOd_LIHT

‘For there fire gives no light’ (royalkgat.tag)

Here the tag for the verbal negator NE signals the presence later in the clause of the negative adjective ‘no’ by the addition in parentheses of the directional arrow > and the usual grammel for adjective, aj.
The flag in parenthesis signals only the part of speech of the additional negative word, not its function in the clause, unless it forms part of the subject of the clause; then it is given the extra flag N (for nominative):

$weila:/int_*WELLE
$lord/n-voc_LOUERD
$/TN_yE
$doel/n_DEOL
$of/pr_OF
$/T<pr_yE
$father/n<pr_FADER
$great/aj-cpv_GRETT+ORE $-er/xs-cpv_+ORE
$never/av_NEUERE
$none/pn_NON
$benot/vpt13{rh}_N+AS $n-/xp-neg(<av<pnN)_N+

‘Alas, lord, the grief of the father — none was ever greater’ (corp145selt.tag)

The tag of the verbal negator is $n-/xp-neg(<av<pnN)_N+, where $n-/xp-neg tells us it is of the type preverbal ‘ne’ cliticised to the verb (in this case the past tense 3rd sg of the verb ‘to be’), and within the following parentheses <av points back to the adverb NEUERE and <pnN to pronominal subject of the clause NON.
As the further examples below illustrate, the additions to the tags are kept as transparent as possible and as similar as possible to tags already in use. The presence of further negative words is signalled by the following labels added to the tag of the verbal negator: adverb = av, adjective = aj, noun = n, pronoun = pn, coordinating sequence = neithernor, and the negative conjunction = nor. The labels aj, n and pn are unmarked when their referents are in oblique cases, but they carry the extra label N (as shown in the example above) when their referent is the subject or part of the subject of the clause. Each additional label is preceded by a directional arrow, indicating whether it occurs before or after the negator. When additional negative words occur with braced negation, their labels are added to the element of the braced negator that occurs on the same side of the verb as they occur. The pronoun ‘naught’, and also the word ‘not’ when it functions not as a verbal negator but as a negative adverb, are treated separately as described in §4.4.5.5 below. Other ‘special’ labels will be described as they occur in the illustrations. More detailed description of the treatment of the conjunction ‘nor’ is also given separately in §4.4.5.6 below.
The remaining examples in this section illustrate some of the variety observable in multiple negation and show how the additional labels in the tags treat it.

$/neg>neg>=>v_*NE
$/P21N_wE
$/neg-v<neg<=(>aj>nor>aj)_NE
$be/vps21_BEOd
$born/vSpp-pl_I+BOR+EN $ge-/xp-vpp_I+ $/vSpp[R]-pl_+EN
$forto/im+H_FOR-TO
$have/vi-m_HABB+ENE $/vi-m_+ENE
$no/ajOd+C_NANE
$pride/nOd_PRUDU
$nor/cj_NE
$forYon/av_FORyE
$no/ajplOd+V_NA\N+E $/plajOd_+E
$other/ajplOd_OdR+E $/plajOd_+E
$renc/nplOd_RENC+AS $/plnOd_+AS

‘We are not born to have any pride or even any other vanities’ (lamhomA1t.tag)

In this example the sequence of additional negative words is two adjectives with the phrasal conjunction ‘nor’ intervening between them. The two adjectives are therefore noted separately. Contrast the example below:

$witannot/vps23-bpn_*N+ETEy $n-/xp-neg(>av2>aj)_*N+
$/P23N_HI
$never/av_NEURE
$whether/pn_wHAyER
$/P23Oi_HAM
$do/vps13_DO+y $/vps13[V]_+y
$worse/av_wRS
$to/pr_TO
$never/av_NEUERE
$no/aj<pr+V_NONE
$gewis/n<pr{rh}_Y+wISSE $ge-/xp-n_Y+

‘They never know (ever to any certainty) which of the two [heat or cold] does them worse harm’ (fmcpmt.tag)

The label (>av2>aj) refers to the sequence NEURE ... NEUERE NONE. The figure 2 attached to the label av makes it clear that in the sequence of additional negative words there are two negative adverbs with no other negative word intervening. In this example, both adverbs happen to be the word ‘never’, but the label av2 would equally be applied if the adverbs were different. The conjunction ‘nor’ more frequently appears in such sequences:

$though/cj_yAH
{>}
$/P13NI_HIT
{>}
$benot/vsjpt13_N+ERE $n-/xp-neg(>av>nor2)_N+
$never/av_NEAUer
$for/pr_FOR
$god/nG_GOD+ES $/Gn_+ES
$love/n<pr_LUUE
{.}
$nor/cj_NE
$for/pr_FOR
$hope/n<pr_HOPE
$of/pr_OF
{\}
$heaven/n<pr_HEOUENE
{.}
$nor/cj_NE
$for/pr_FOR
$dread/n<pr_DRED
$of/pr_OF
$hell/n<pr_HELLE

‘... though it were never for God’s love, nor for hope of heaven, nor for fear of hell...’ (bod34t.tag)

No differentiation is made in the sequential labelling of ‘nor’ between those examples linking different kinds of subsidiary phrasal elements (though see below §4.4.5.6 for the treatment of ‘nor’ linking different main clauses):

$all/ajnpl_*ALL+E $/plajn_+E
...
{\}
$/neg-v(>nor2)_*NE
$may/vpt23_MIzTE
$tell/vi_TELL+E $/vi_+E
$nor/cj_NE
$in{p}/pr_IN
$book/n<pr_BOKE
$read/vi{rh}_RED+E $/vi{rh}_+E
{\}
$all/ajOd_*ALLE
$/TOd_yE
$sore/nOd_SORE
$nor/cj_NE
$all/ajOd_ALLE
$/TOd_yE
$dread/nOd{rh}_DREDE

‘All [of them] could not describe, nor read in a book, all the pain nor all the dread...’ (digby86mapt.tag)

Here the label (>nor2) simply draws attention to the presence of the two following negative conjunctions, though one links the infinitives TELLE and REDE, while the other links the two predicate phrases dependent on them. The extended tagging provides preliminary notes for whatever more detailed syntactic analysis may be desired. Very often such linked phrases are made contrastive by the presence of a leading ‘neither’:

$that/cj_y~
{\}
$/P22Oi_OU
$/neg-v(>neithernor)_NE
$lystan/vsjps13_LUSTE
$neither/cj>=_NOUdER
$speak/vi_SPEOK+EN $/vi_+EN
$nor/cj<=_NE
$hearken/viK2_HERCHN+IN $/viK2_+IN
{\}
$worldly/ajOd_wORLD+LICHE $-ly/xs-ajOd_+LICHE
$speech/nOd_SPECHE

‘...that it should please you neither to speak nor to listen to wordly speech’ (cleoarat.tag)

A leading ‘neither’ may introduce a string of following ‘nor’ conjunctions; the label in the negator’s tag is then given the appropriate number (cf. §4.4.8.2 below):

$2fold/qc-aj_*TwI+FOLD $-fold/xs-aj_+FOLD
$or/cj_OdER
$manifold/aj_MANI+FOLD $-fold/xs-aj_+FOLD
$be/vps13_IS
$/TN_yE
$man/n_MAN
{.}
$/RTA_yE
$benot/vps13_N+IS $n-/xp-neg(>neithernor4)_N+
$steadfast/aj_STEDEFAST
$neither/cj>=_NE
$in{m}/pr_ON
$deed/n<pr_DADE
{.}
$nor/cj<=_NE
$of/pr_OS
{=sic=}
$speech/n<pr_SPECHE
{.}
$nor/cj_NE
$in{m}/pr_ON
$thank{c}/n<pr_yONKE
{.}
$nor/cj_NE
$against{t}/pr_GENES
$god/n<pr_GODE
{.}
$nor/cj_NE
$against{t}/pr_TO-GENES
$man/n<pr_MAN

‘Twofold or manifold is the man who is steadfast neither in deed, nor of speech, nor in thought, nor towards God, nor towards man’ (trhomAt.tag)

Here the label (>neithernor4) indicates a following coordinating ‘neither + nor’ construction with three more ‘nor’ conjunctions thereafter. Consider this example:

$nor/cj_NE
$never/av_NEAUER
$no/aj+C_NA
$nauDsyn/n_NOwCIN
$/neg-v(<norC<av<ajN>neither3nor3>nor>ajN)_NE
{\}
$eglian/vps13K2_EIL+ET $/vps13[L]K2_+ET
$there/av_yER
$man/nOd_MON
$neither/cj>=_NOwdER
$sorrow/n_SORHE
$nor/cj<=_NE
$sore/n_SAR
$neither/cj>=_NOwdER
$heat/n_HEA\TE
$nor/cj<=_NE
$cold/n_CHELE
$neither/cj>=_NOwdER
$hunger/n_HUNGER
{<NE CHELE<}
{=del, repeated in error and erased=}
$nor/cj<=_NE
$thirst/n_yURST
{\}
$nor/cj_NE
$no/aj+V_NAN
$ofYyncan/vn_OF-yUNCH+UNGE $of-/xp-v_OF- $/vn_+UNGE

‘...nor does ever any distress there annoy a man, neither sorrow nor pain, neither heat nor cold, neither hunger nor thirst nor any displeasure’ (royalkgat.tag).

Here the label >neither3>nor3 makes it clear that there are three of each of ‘neither’ and ‘nor’, not just ‘neither’ plus three following ‘nor’, which would be implied by the label >neithernor3. For the label <norC, see §4.4.5.6 below. Note that occasional the combination ‘either + or’ is used within the scope of multiple negation. This is simply labelled as such in the tag of the verbal negator as usual:

$/DatpnRTA_*yE-yE
$/neg-v(>eitheror+neg)_NE
$can/vps13_CON
$either/cj>=_OdER
$u:htsong/nOd_UHT+SONG $song/nOd-k_+SONG
$or/cj<=_OyER
$/neg-v_NE
$may/vps13_MEI
$/P13OdI_HIT
$say/vi_SIGG+EN $/vi_+EN
$for/pr_UOR
$u:htsong/n<pr_UHT+SONG $song/n<pr-k_+SONG
$say/vsjps13_SIGGE
$30/qc_yRITTI
{(PATER NOSTERES(}

‘She who does not know matins, or cannot sing it, instead of matins let her sing thirty Pater Nosters’ (neroart.tag)

Here there are two negative clauses syntactically overlapping, and this is recognised in the label (>eitheror+neg) where neg refers to the negator of the second negative clause. More complex or extended negative constructions are dealt with on a similarly ad hoc basis, but the additions to the tags are usually transparent in context. One such example is given below:

$benot/vps13_*N+IS $n-/xp-neg(>pnN>negsubcl+nor+naughtN3>negsubcl)_*N+
$none/pn_NON
{=A later hand adds T to form NOUT=}
$so/av_SO
$hot/aj_HOT
{.}
$that/cj_yer
$/P13NI_HIT
$/neg-v_N+
$acool/vps13K2{rh}_+A-COL+Ey $a-/xp-v_A- $/vps13[L]K2{rh}_+Ey
{.}
{\}
$nor/cj_*NE
$naught/n_NOzT
$so/av_SO
$white/aj_HwIT
{.}
$that/cj_yer
$/P13NI_HIT
$/neg-v_NE
$solian/vps13K2{rh}_SOL+Ey $/vps13[L]K2{rh}_+Ey
{.}
{\}
$nor/cj_*NE
$naught/n_NOzT
$so/av_SO
$le:of/aj_LEOF
{.}
$that/cj_yer
$/P13NI_HIT
$/neg-v_NE
$ala:Yian/vps13K2{rh}_A+LOy+Ey $a-/xp-v_A+ $/vps13K2{rh}_+Ey
{.}
{\}
$nor/cj_*NE
$naught/n_NOzT
$so/av_SO
$glad/aj_GLAD
{.}
$that/cj_yer
$/P13NI_HIT
$/neg-v_NE
$awrath/vps13K2{rh}_A-wROy+Ey $a-/xp-v_A- $/vps13K2{rh}_+Ey

‘There is nothing so hot that it doesn’t cool down, nor naught so white that it doesn’t get dirty, nor naught so dear that it doesn’t become hateful, nor naught so delightful that it doesn’t become annoying’ (cotowlbt.tag)

All the negative subordinate clauses are dependent on the original NIS (‘there is not’) so all the negative words may be seen as examples of multiple negation within the scope of the negator of the finite verb. As the structure is repeated, the most economical way to signal its complexity is with the tag: $n-/xp-neg(>pnN>negsubcl+nor+naughtN3>negsubcl). $n-/xp-neg shows that the negator of the main clause is a cliticised ‘ne’. The first pronoun (NON) is marked as the first subject with the label >pnN. Thereafter >negsubcl+nor+naughtN3>negsubcl indicates that there is a negative subordinate clause followed by ‘nor’ and subject ‘naught’ (naughtN), the + signs and figure 3 indicating that the sequence linked by the + signs occurs three times. There is then a final negative subordinate clause.

§4.4.5.4 The negative exclusion construction – ‘ne + but’55
Middle English ‘ne + but’ is the precursor of the late Middle English and modern English use of ‘but’ with the sense ‘only’ — ‘I am but a poor boy’. The emergence of this ‘exclusive “but”’ parallels the reduction of ‘ne + not’ to post-verbal ‘not’ alone in Jespersen’s cycle. In the LAEME tagging system we treat ‘ne + but’ as part of the early Middle English system of verbal negation, and signal the presence of ‘but’ in negative exclusion constructions by a label added to the negator’s tag:

$/neg-v(>but)_NE
$have/vps12_HAW+YST $/vps12_+YST
$/P12N_yU
$here/av_HERe
$bu:tan{o}/av_BOTE
$fight/nOd{rh}_FYT

‘You have here only struggle’ (adde6at.tag)

We translate this as it would be put in modern English and the BOTE is given the tag $bu:tan{o}/av, where the {o} flag indicates the sense ‘only’. But because the early Middle English construction is still overtly negative, this example might also be translated: ‘You do not have here other than struggle’ which shows that at there is a contrastive, as well as an exclusive, semantic element. The construction ‘ne + but’ may also combine with other negative words, in which case the contrastive force is extended and it becomes semantically possible to substitute ‘except’ for ‘only’ or ‘other than’ in the modern translation:

$maiden/n_*MAIDEN
$&/cj_AND
$mother/n_MODER
$benot/vpt13_N+AS $n-/xp-neg(>av>ajN>but)_N+
$never/av_NEUer
$no/aj+C_NON
{=erasure=}
$woman/n_WIMON
$bu:tan{e}/av_BOTEn
{\}
$/P13NF{rh}_HE

‘Maiden and mother, was never any woman except she’ (tr323ct.tag)

Here BOTEn is given the tag $bu:tan{e}/av, where the {e} flag indicates the contrastive sense ‘except’.
As with other negative words in multiple negation constructions, ‘but’ resists combination with braced negation of the ‘ne + not’ type, single ‘ne’ negation being preferred in combination with exclusive ‘but’. Indeed Laing (2002: 309) expressed the opinion that there were no early Middle English exceptions to this preference since in any apparent exceptions one could analyse the ‘not’ element of the braced negation as ‘naught’ with the sense ‘nothing’ forming with following ‘but’ a prepositional phrase of exception — ‘nothing except’. In fact there are a few examples in the LAEME corpus where it is not sensible to analyse them in any other way than as braced negation + ‘but’, so some of the potentially ambiguous ones have also been accepted as such in the tagging. They are, however, very few. One unambiguous one and one ambiguous one are given below:

$/RTA_yET
$/neg-v>=_NE
$can/vps13_KAN
{,}
$not/neg-v<=(>but)_NAzT
{\}
$sing/vi_ZING+E $/vi_+E
{,}
$bu:tan{e}/av_BOTE
$of/pr_OF
$/P13<prXM_HIm-ZELUE $-self/xs-<pr_-ZELUE

‘Who cannot sing except/other than about himself’ (ayenbitet.tag)

The word order makes it clear that the negation here is braced verbal negation and it would unduly force the syntax to read ‘Who can sing nothing except about himself’. The next example is less clear:

$/P13GM_HIS
$thirst/n_yURST
$benot/vps13_N+IS $n-/xp-neg>=_N+
$not/neg-v<=(>but)_NOUT
$bu:tan{o}/av_BUTEN
$geornan/vn_gIRN+UNGE $/vn_+UNGE
{\}
$of/pr_OF
$/P21G_URE
$soul/nG_SOUL+E $/Gn_+E
$hae:lu/n<pr_HELE

‘His thirst is only yearning of our soul’s healing’ (neroart.tag)

Note that this could be tagged as follows:

$/P13GM_HIS
$thirst/n_yURST
$benot/vps13_N+IS $n-/xp-neg(naughtN>but)_N+
$naught/n_NOUT
$bu:tan{e}/av_BUTEN
$geornan/vn_gIRN+UNGE $/vn_+UNGE
{\}
$of/pr_OF
$/P21G_URE
$soul/nG_SOUL+E $/Gn_+E
$hae:lu/n<pr_HELE

‘His thirst is nothing except yearning of our soul’s healing’

But (pace Laing 2002) given the existence of some unequivocal exceptions to Jack’s Law in this construction, and also that the other versions of the text (Ancrene Riwle) here have ‘ne + but’ rather than ‘ne + not + but’, it seems reasonable to tag it as we now have as braced negation plus ‘but’. However, the presence of such ambiguities in the reading of constructions that include ‘not/naught’ (< OE )(n)wiht means that it is useful to label examples of these words separately in the negation tags.

4.4.5.5 Treatment of ‘not’ and ‘naught’
Historically, ‘naught’ and the verbal negator ‘not’ are the same word: both originate from OE )(n)wiht, nōht, the negated form of ā)wiht ‘anything’. By the early Middle English period the functions of ‘not’ and ‘naught’ are usually distinguishable, but the spellings within individual text languages are not differentiated according to function. Presumably it was felt that the two functions were still carried out by ‘the same word’ although the functions themselves were for the most part distinguished. This does, however, make for potentially ambiguous cases, especially when ‘naught’ is used adverbially, so in the special labels to the negator tags, we differentiate the word from other negative adverbs, both when it functions as ‘not’ and when it functions as ‘naught’.

4.4.5.5.1 Treatment of ‘not’
We have already described the tags for ‘not’ when it forms all or part of the negator of the verb in clausal negation. There are tags assigned to adverbial ‘not’ that need to be explained in relation to clausal negation. The tag $not/av is given to the word ‘not’ when nothing is negated except the word or phrase immediately in the domain of the ‘not’ and where no clausal negation is implied:

$&/cj_*AND
$draw/vSpt23_DROW+En $/vSpt23_+En
$/P13OdM_HIM
$unto/pr_UN-TO
$/Tpl<pr_yE
$gallows/npl<pr{rh}_GALW+ES $/pln<pr{rh}_+ES
{\}
$not/av_*NOUTH
$by/pr_BI
$/T<pr_yE
$gate{w}/n<pr_GATE
$but/cj_*BUT
$over/pr_OUER
$/Tpl<pr_yE
$fallow/npl<pr{rh}_FALW+ES $/pln<pr{rh}_+ES

‘And [they] dragged him to the gallows, not by the gate but over the fields’ (havelokt.tag)

The same tag is used for ‘not’ when it does appear in a negated clause but when it serves merely as a negative adverb and when the function of verbal negator is fulfilled by another word or words. In these cases, however, the tag on the verbal negator signals that the adverb is a form of ‘not’ by having the special label notav. In the following example, compare the tag labels >av, referring to NO and >notav referring to NOUzT:

$here/av_*HERE
$willnot/vps11_N+ELLE $n-/xp-neg(>av>nor>notav)_N+
$/P11N_ICH
$no/av_NO
$long/av-cpv_LENG
{=G changed from D. No sign of the abbreviation for <er> which}
{is supplied by Horstmann (1887)=}
$dwell/vi{rh}_DWELL+E $/vi{rh}_+E
{\}
$nor/cj_*NE
$not/av_NOUzT
$more/av_MORE
$with{w}/pr_WITH
$/P12<pr_yE
$spellian/viK2{rh}_SPELL+E $/viK2{rh}_+E

‘Here will I no longer dwell, nor speak with thee any more’ (laud108at.tag)

All examples of the adverb ‘not’ used within clausal negation may thus be easily identified and their contexts isolated for further syntactic analysis.
There are some examples of adverbial ‘not’ that seem to have a closer link to the negation of the verb than the examples above, without actually having the function of primary clausal negator. These are given the tags $not/av<v and $not/av>v. The tags are used for adverbial ‘not’ when the finite verb governs a positive clause, but when subsequent ‘not’ implies an understood repetition of the verb of the positive clause (either before or after it — hence the directional arrow) which produces an alternative negated clause.

$in{b}/pr_ON
$/P12<pr_yE
$/P11N_I+
$trust/vps11_+TRUSTE
$not/av<v_NAwT
$in{b}/pr_O+
$/P11<prX_+ME-SEOLUEN $-self/xs-<pr_-SEOLUEN

‘I trust in thee not in myself’ (royalkgbt.tag)

This has the implied underlying structure: ‘I trust in thee [I trust] not in myself’. The example below illustrates the same phenomenon but with ‘not’ occurring before the verb:

$for/pr_*FOR
$/P21G_HORE
$sin/npl<pr_SUNN+EN $/pln<pr_+EN
$&/cj_AND
$not/av>v_NOUT
$for/pr_*FOR
$/P13GMplpn<pr_HIS
$/P13GM_HIS
$flesh/nOd_FLES
$/P13NM_HE
$sell/vpt13{rh}_SOL+DE $/vpt13[L]{rh}_+DE

‘For our sins, and not for his, he sold his flesh’ (digby86bodysoult.tag)

Here the underlying structure is: ‘He sold his flesh for our sins and [he sold his flesh] not for his [sins]’. In modern English we might express this as: He did not sell his flesh for his own sins but for ours’. Of course, the semantics of these constructions is in each case positive, not negative: ‘I trust’, ‘he sold’. The presence in the tag of <v or >v merely draws attention to the formal possibility of analysing these examples of ‘not’ in terms of clausal negation.
Because Middle English (like modern English) uses clausal negation even when the scope of the negation is not the main clause, this use of ‘not’, with only implied verbal negation, is given the same tag also in cases where the scope of the negation is outside the main clause. The usage is thereby flagged for more detailed analysis:

$/RTI_y~
$be/vpt13_wES
$wunian/vppK2_I+wUN+ET $ge-/xp-vpp_I+ $/vpp[N]K2_+ET
$to/im+C_TO
...
$do/vi_DO+N $/vi[V]_+N
$after/pr_EFTE*R
$will/n<pr_wIL
$/P23G_HA*RE
$lady/n<pr_LEFDI
$&/cj_*ANT
{\}
$not/av<v_NAwT
$after/pr_EFTE*R
$wit/n<pr_wIT

‘Which [the household] was accustomed to ... act according to Will their lady and [was accustomed to act] not according to Wit’ (royalkgbt.tag)

Such implied negations may also have added negative words, which are labelled in the tagging in the usual way:

$herein{m}/av-k_*HER-IN $in{m}/pr-k_-IN
$be/vps13_IS
$religion/n_RELIGIUN
$not/av<v(>nor3)_NAUT
$in{m}/pr_I+
$/T<pr_+yE
{\}
$wide/aj<pr_wIDE
$hood/n<pr_HOD
{.}
$nor/cj_NE
$in{m}/pr_I+
$/T<pr_+yE
$black/aj<pr_BLAKE
{<}
{=del, by erasure=}
{.}
{<}
$nor/cj_NE
$in{m}/pr_I+
$/T<pr_+dE
$white/aj<pr_HwITE
$nor/cj_NE
{\}
$in{m}/pr_IN
$/T<pr_yE
$grey/aj<pr_GREIzE
$cowl/n<pr_CUUEL

‘Herein is religion, [religion is] not in the wide hood, nor in the black, nor in the white, nor in the grey cowl’ (cleoarat.tag)

The same tag is also used to indicate ‘not’ in the second half of a braced infinitive clause whether the main clause is negated or not:

$/RTApl_yET
$/neg-v>=_NE
$may/vps23_MOzE
$/P23OdX_HAm
$not/neg-v<=(>nor>notav)_NAzT
$hold/vi_HYALD+E $/vi_+E
$still/av_STILLE
{,}
$nor/cj_NE
$not/av<v_NAzT
$/P23OdX_HAm
$look{p}/viK2_LOK+I $/viK2_+I

‘... who may not hold themselves still nor [may not] protect themselves’ (ayenbitet.tag)

$/P13NM_HE
$/P13OdXM_HIm
$shall/vps13_SSEL
$much/av_MOCHE
$mildian/viK2_MILD+I $/viK2_+I
{"humble"}
{.'}
$&/cj_AND
$not/av<v_NAzT
$/P13OdXM_HIm
$praise/viF_PRAYS+Y $/viF_+Y

‘He shall greatly humble himself, and [shall] not praise himself’ (ayenbitet.tag)

4.4.5.5.2 Treatment of ‘naught’
The tag $naught/n is given to the early Middle English variants of the reflex of nā(n)wiht/nōht when the word is a negated noun meaning ‘nothing’. The word may also be used adverbially meaning ‘not at all’ ($naught/av) and occasionally as an adjective meaning ‘worthless’ ($naught/aj). When ‘naught’ in any of these functions occurs within the scope of the negator of a negative clause, its presence is signalled in the additional labels to the negator’s tag by using a form of its full tag so as to differentiate it from other negative nouns, adjectives and adverbs:

$for/cj_*FOR
$man/indef_ME
$/neg-v(>naught)_NE
$shall/vps13_SSEL
$naught/nOd_NOzT
$for/pr_FOR
$/P13<prM_HIm
$do/vi_DO $/vi[V]_0

‘For one shall do nothing for him’ (corp145selt.tag)

$benot/vsjpt13_N+ERE $n-/xp-neg(>naughtN)_N+
$/P23Oi_HOM
$naught/n_NOzT
$so/av_SO
$loth/aj{rh}_LOyE

‘There should be nothing so hateful to them’ (corp145selt.tag)

$/P12N_yU
$benot/vps12_N+ART $n-/xp-neg(>naughtajN>but)_N+
$naught/aj_NOUHT
$bu:tan{e}/av_BUTE
$dead/aj{rh}_DED

‘Thou art worthless except dead’ (jes29t.tag)

$/neg-v(>aj>naughtav)_*NE
$have/vps13_HAU+Ed $/vps13_+Ed
$/DatN_dAT
$venom/n_UENIM
$no/ajOd+C_NOn
$might/nOd{rh}_MIGT
{.}
$to/im+C_TO
$derian/viK2[1]-m_DER+EN $/viK2[1]-m_+EN
$/P13OdM_HIM
{\}
$since/av_SIdEN
$naught/av{rh}_NOn-wIGT

‘That poison has no power to harm him afterwards in any way’ (bestiaryt.tag)

4.4.5.5.3 Adverbial ‘naught’ and braced ‘ne + not’ negation
The texts in the LAEME corpus show no significant difference in their variant spellings of the reflexes of OE nā(n)wiht/nōht between their use as ‘naught’ and as ‘not’ in any particular scribe’s usage. Although they are sometimes resolvable by word order, ambiguities can therefore arise between ‘naught’ used as an emphatic adverb meaning ‘not at all’ and ‘not’ as the second part of braced ‘ne + not’ negation. Consider the following two examples:

$/TN_*yE
$rich/aj_RICHE
$earl/n_ERL
$/neg-v>=_NE
$forget/vSpt13_FOR+YAT $for-/xp-v_FOR+
$not/neg-v<={rh}_NOUTH

‘The powerful earl did not forget’ (havelokt.tag)

$for/cj_*FOR
$/P11N_I+
$/neg-v(>av>naughtav)_+NE
$misdo/vpt11_MIS+DEDE $mis-/xp-v_MIS+
$/P13OiM_HIM
$never/av_NEUerE
$naught/av{rh}_NOUTH

‘For I never mistreated him in any way at all’ (havelokt.tag)

There seems to be a clear semantic split between ‘not’ in the first example and ‘naught’ in the second although both are spelled NOUTH. But now consider the next two examples:

$have/v-imp_*HAUE
$/P12N_yOU
$not/neg-v<(>notav)_NOUTH
$thereof/av-k_yER-OFFE $of/pr-k_-OFFE
$doubt/nOd{rh}_DOUTHE
{\}
$not/av_*NOUTH
$/T-av_yE
$worth/n-av_WORTH
$of/pr_OF
$/A<pr_ONE
$nut/n<pr{rh}_NOUTHE

‘Do not have any doubt about it, not so much as a nut’s worth’ (havelokt.tag)

Here the clausal negator is post-verbal ‘not’ alone. The following negative adverb seems not to have any special emphasis so has been tagged as $not/av and signalled with the label (>notav) in the negator’s grammel. But the next example is more difficult to tag:

$/TN_*yE
$man/n_MAn
$/RTA_yAT
$may/vps13_MAY
$well{w}/av_WEL
$eat/vi_ET+En $/vi_+En
$&/cj_AND
$drink/vi{rh}_DRInK+En $/vi{rh}_+En
{\}
$/DatpnOd_*yAT
$not/neg>v>=_NOUTH
$/neg>v<=(>but)_NE
$have/vsjps13_HAUE
$bu:tan{e}/av_BUT
$on{m}/pr_ON
$swinc/n<pr_SWInK
$along/av{rh}_LOnG

‘The man who is able to eat and drink well, he may not achieve that except by means of work’ (havelokt.tag)

We have chosen to take this structure as an example of braced preverbal ‘not + ne’ negation, not least because there are six other examples of this kind of negation in the text (Havelok). But it could in principle be tagged like this:

$/DatpnOd_*yAT
$naught/av_NOUTH
$/neg-v(<naughtav>but)_NE
$have/vsjps13_HAUE
bu:tan{e}/av_BUT

‘He may not at all achieve that except...’

or indeed, as in the previous example, taking NOUTH as a separate adverbial ‘not’:

$/DatpnOd_*yAT
$not/av_NOUTH
$/neg-v(<notav>but)_NE
$have/vsjps13_HAUE
bu:tan{e}/av_BUT

‘He may not [not] achieve that except...’

One consideration that somewhat militates against the tagging we have in fact chosen is that it thus constitutes an exception to Laing’s corollary to Jack’s Law.

4.4.5.5.4 Corollary to Jack’s Law and the appearance of adverbial ‘nothing’ and ‘none’
There is not only a strong resistance in early Middle English texts to combining ‘not’ in braced negative constructions with other negative words; this resistance to negative concord seems to be evinced also by ‘naught’. This observation is expressed by Laing (2002: 305) as a corollary to Jack’s Law, where the corollary is added to the original law in italics: ‘ne..nawt is not normally used in clauses containing a further negative form, with exception of the conjunction ne, whether or not nawt retains the full sense of ‘not at all’ or even when nawt is a negated noun meaning ‘nothing’. Laing suggests that ‘when non-emphatic, grammaticalised ‘ne..not’ became increasingly widespread, the formally identical ‘ne..naught’ construction became subject to the same syntactic restraints’. This conclusion is supported by appearance of the adverbial use of the negative noun ‘nothing’, instead of adverbial ‘naught’, in constructions that could otherwise be ambiguous. In order to facilitate comparison, this adverbial ‘nothing’ is signalled with a distinct label (nothingav) in the tag of the verbal negator:

'_*IOSEP
$of/pr_OF
$/Des<pr_yESE
$word/npl<pr_WORD+ES $/pln<pr_+ES
$benot/vpt13_N+AS $n-/xp-neg(>nothingav)_N+
$/P13NM_HE
$nothing/n-av_NO-yING $thing/n-av-k_-yING
$glad/aj{rh}_GLAD

‘Joseph was not at all delighted at these words’ (iacobt.tag)

$forth/av_FORy
$/P23N_HI
$row/vSpt23_REW+E $/vSpt23_+E
$fast/av{rh}_VASTE
{\}
$&/cj_*AND
$trust/vpt23_Tr^IST+E $/vpt23[D]_+E
$all/av_AL+
$to/pr+V_+TO
$/P21G_OURe
$lord/nG_LOUerD+ES $/Gn_+ES
$grace/n<pr_Gr^ACE
{.}
$&/cj_&
$nothing/n-av_NO+yInG $thing/n-av-k_+yInG
$benot/vpt23_N+ERE $n-/xp-neg(<nothingav)_N+
$agasten/vpp-pl{rh}_A+GASTE $a-/xp-v_A+

‘Forth they rowed quickly and trusted entirely to Our Lord’s grace and were not at all frightened’ (corp145selt.tag)

The word ‘none’ is also occasionally used adverbially with the sense ‘not [at all]’. This is most often in phrases with ‘none the’ plus comparative (‘none the better’) or ‘none so’ plus adjective (‘none so great a truth’). But in the LAEME corpus we have found two examples, both in northern texts, where it seems to function, like ‘nothing’ above, as an alternative to emphatic ‘naught’. For comparison therefore, all examples of adverbial ‘none’ when they are found in negative clauses are signalled in the verbal negator’s tag with the label noneav:

$when/cj_*QUEN
$/P13NM_HE
$be/vpt13_WAS
$dead/aj_DED
$/neg-v(>noneav)_NE
$doubt/vpt13F_DOUT $/vpt13F_0
$/P13NM_HE
$none/av{rh}_NAN
{\}
$that/cj_Y^T
$/P13GM_HIS
$soul/n_SAUL
$/neg-v_NE
$be/vpt13_WAS
$to/pr+H_TIL
$hell/n<pr_HEL
$go/vSpp{rh}_GA+N $/vSpp[V]{rh}_+N

‘When he [Adam] was dead he doubted not at all that his soul would go to to hell’ (cotvespcmat.tag)

$none/av_*NAN
$may/vpt13_MIHT
$/P13NM_HE
$not/neg-v<(<noneav)_NOHT
$with{w}/pr_WIT
$/P13GM_HIS
$main/n<pr{rh}_MAIn
{\}
$ilk/DatOd_*yAT-ILK
$earth/nOd_ERyE
$make/viK2_MAK $/viK2_0
$flesh/nOd_FLEIS
$again/av{rh}_OGAIn

He might not in any way, by his own strength, turn that earth back into flesh’ (edincmat.tag)

4.4.5.6 Treatment of ‘nor’
In the examples above, presence of the conjunction ‘nor’ is signalled in the negator’s tag with the label <nor and >nor, when ‘nor’ is a phrasal conjunction linking words or phrases within the scope of the negaotor of the negative main clause. When it appears in the coordinating structure ‘neither..nor’, the presence of the combined structure is signalled by the labels <neithernor and >neithernor. In these contexts, ‘nor’ forms part of multiple negation and its presence is therefore indicated alongside other negative words within the scope of the negator of the negative clause. When ‘nor’ is functioning as a clausal conjunction, however, it is not part of multiple negation because it lies outside the negative clause itself. We have no reason to suppose that the different syntactic functions of ‘nor’ affect its spellings or their regional distributions. So as dialectologists we have no special reason to separate the two usages in the tagging system. Both are therefore given the tag $nor/cj and all spelling variants of all functions of ‘nor’ are listed together in the text dictionary of each tagged text and will appear together when mapped. But it is clear that in the context of the history of verbal negation the distinction between ‘nor’ functioning as a phrasal conjunction or as a clausal conjunction is important. The strength of the tagging system as it stands is that it forms a basis for possible further syntactic parsing of the tags themselves in order to isolate different syntactic functions. Meanwhile we have gone some way towards this by giving a special label (norC) in the additions to the negators’ tags to differentiate ‘nor’ as a clausal conjunction linking separate negative clauses from ‘nor’ as a phrasal conjunction operating within multiple negation.
The conjunction ‘nor’ has five possible lexels: $&nor/, $nor/, $nor&/ and $nor[neg]/ $nor[or]/. $&nor/cj is assigned to the text-form NE when it is used as a conjunction linking two positive clauses, where in modern English the word ‘and’ would always be used. Similarly, $nor[or]/cj is assigned to the text-form NE when it links two positive alternatives in exactly the same way as a form of the word ‘or’. These uses are very rare and do not come into the realm of verbal negation. $nor/cj is assigned to forms of the word ‘nor’ when they link two or more negative structures, including negative clauses. In these cases, the verbal negators in each negative clause will carry a label and directional arrows signalling its presence and its function as a clausal conjunction outside the scope of the clause itself:

$/neg-v(>aj>norC)_*NE
$reccan/vps13_RECCH+Ed $/vps13_+Ed
$christ/n_Cr^IST
{\}
$no/ajOd+C_NANE
$le:asung/vnOd_LEAS+UNGE $/vnOd_+UNGE
$nor/cj_NE
$/P13OiM_HIM
$benot/vps13_N+IS $n-/xp-neg(<norC>ajN)_N+
$no/aj+C_NA
$need/n_NEOd

‘Christ takes no heed of any lying, nor does he have any need of it’ (lamhomA1t.tag)

When there is a sequence of negative clauses with braced negation, as usual, the norC label is placed in the tag of the negator nearest to the conjunction:

$ac/cj_*AH
$all/ajnpl_ALL+E $/plajn_+E
$/neg-v>=_NE
$may/vps23_MAHE
{\}
$not/neg-v<=(>norC)_NAwT
$hold/vi_HALD+EN $/vi_+EN
$1/qcOd_A
$rule/nOd_RIwLE
{.}
$nor/cj_NE
$/neg-v(<norC)>=_NE
$Yurfan/vps23_yURUE
$not/neg-v<=(>norC)_NAwT
$nor/cj_NE
$/neg-v(<norC)>=_NE
$a:gan/vps23_AHE
$not/neg-v<=_NAwT
$hold/vi_HALD+EN $/vi_+EN
$in{m}/pr_ON
$1/qc<pr_A
$wi:se/n<pr_wISE
$/TOd_yE
$outer/ajOd-cpv_UTT+RE $-er/xs-cpv_+RE
$rule/nOd_RIwLE

‘But not everyone may hold a rule, nor need they, nor ought they hold the outer rule in one way’ (corpart.tag)

The tag $nor&/cj is assigned to forms of ‘nor’ when it links a positive clause with a following negative one where in modern English we would normally use the word ‘and’ (note also the use of ‘or’, thus labelled, where we would normally expect ‘nor’ with multiple negation:

$&/cj_AND
$together/av_TO-GYDERE
$/P23N_HY
$be/vps23-apn_BYEy
$pi:nian/vpp-plK2_Y-PYN+ED $ge-/xp-vpp_Y- $/vpp[N]-plK2_+ED
{.}
$nor&/cj_NE
$never/av_NEURE
$/neg-v(<nor&C<av>or)_NE
$shall/vps13_SSEL
$be/vi_BY $/vi[V]_0
$end/n_ENDE
$of/pr_OF
$pi:ne/n<pr_PYNE
{.'}
$or/cj_OyER
$rest/n_RESTE

‘And together they are tormented; and there shall never be and end of the torment nor any rest’ (ayenbitet.tag)

The tag $nor[neg]/cj is assigned to forms of ‘nor’ when it functions also as (or instead of) a verbal negator for the following negative clause:

$stand/v-imp_*STOND
$nor[neg]/cj_NE
$fall/v-imp_FAL
$nomore/av_NAMORE
$down/av{rh}_ADOUN

‘Stand and do not fall down any more’ (digby86mapt.tag)

$/P13NM_*HE
$/P23Od_HEM
$clothe/vpt13K2_CLOy+EDE $/vpt13K2_+EDE
$right/av_RITH
$nor[neg]/cj_NE
$feed/vpt13{rh}_FEDDE

‘He did not clothe them or feed them properly’ (havelokt.tag)56

This usage is found rarely in the LAEME corpus and we have found only one example where it combines with another negative clause:

$&/cj_&
$be/vsjps21-bpn_BEO
$/P21N_wE
$swi:De/av_SwIyE
$Yolemo:d/ajpl{rh}_yOLEMOD+E $/plaj{rh}_+E
$that/cj_y~
$/P21N_wE
$/TOd_yENNE
$fiend/nOd_FOND
$not/neg>v>=_NOHT
{\}
$/neg>v<=(>nor[neg])_NE
$dread/vsjps21{rh}_DREDEN
$nor[neg]/cj_NE
$/P13NM_HE
$/P21Od_US
$beswi:can/vsjps13_BI+SwIKE $be-/xp-v_BI+
$through/pr_yURH
{=MS yURdH with d subpuncted=}
$/P13GMpl_HIS
$deed/npl<pr{rh}_DED+EN $/pln<pr{rh}_+EN

‘And let us be very patient so that we may not fear the devil, so that he may not deceive us through his deeds’ (lamhomA2t.tag)

Of more frequent occurrence is a structure consisting of two syntactically discrete, but semantically connected, negative clauses where the verbal negator for the second clause may be taken to function also as a clausal conjunction ‘nor’. In these cases the label >neg[nor] is added to the tag of the negator in the first of the two negative clauses:

'_*SATANAS
$here/av_HERE
$/P11N_*I
$/P12Od_yE
$bind/vps11{rh}_BINDE
{\}
$/neg-v(>av>neg[nor])_*NE
$shall/vps12_SALT
$/P12N_yOU
$never/av_NEUERE
$hence/av_HENE
$wind/vi{rh}_WIND+E $/vi{rh}_+E
{\}
$/neg-v(>av)_*NE
$shall/vps12_SALT
$/P12N_yOU
$never/av_NEUERE
$wendan/vi_WEND+EN $/vi_+EN
$away/av{rh}_AWAY
{.}
{\}
$until/cj_*TILLE-yAT
$come/vps13_COM+Ey $/vps13[N]_+Ey
$doomsday/n{rh}_DOMES+DAY $day/n-k{rh}_+DAY

‘Satan: “Here I bind thee. Thou shalt never turn from here, nor shalt thou ever go away, until doomsday comes”’ (digby86mapt.tag)

Note that in the label nor[neg] the brackets indicate that the presence of a verbal negator is implied within the function of the ‘nor’ conjunction. Similarly in the label neg[nor] the brackets indiacate the presence of an implied conjunctive ‘nor’ within the function of the verbal negator.
A structure found commonly in the LAEME corpus is a preverbal ‘ne’ negator plus finite verb followed by ‘nor’ followed by a coordinating finite verb. These constructions are precursors of the type ‘We neither sleep nor rest’ in modern English implying ‘we do not sleep and we do not rest’. These types are signalled by the combination [neither]>norC added to the tag of the negator, where the brackets again indicate that there are two parallel negated verbs, where the first negator functions both as the negation marker and an implied ‘neither’ and the second acts also as a coordinating nor plus verb:

$that/cj_y~
$/P23N_HIE
$/neg-v([neither]>norC_NE
$see/vps23_SI+EN $/vps23[V]_+EN
{*sj context*}
$nor/cj_NE
$understand/vps23_UNDER+STAND+E $under-/xp-v_UNDER+ $/vps23_+E
{*sj context*}

‘... so that they may neither see nor understand’ (vvbt.tag)

When this structure occurs with braced negation, the label coordinating ‘nor’ is signalled in the tags of both elements of the double negation (>norC and <norC) to indicate that coordinates with both:

$that/cj_*yAT
$/P22N_zE
$/neg-v([neither]>norC)>=_NE
$eat/vSpt22_ET+E $/vSpt22_+E
$nor/cj_NE
$drink/vSpt22_DRONK+E $/vSpt22_+E
$not/neg-v<=(<norC>norC)_NOzT
{.}
$nor/cj_NE
$sleep/vSpt22_SLEP+E $/vSpt22_+E
$with{w}/pr_WIy
$/P22G_zOURe
$eye/npl<pr{rh}_EIzE

‘... that ye neither ate nor drank nor slept with your eyes’ (corp145selt.tag)

Contrast the following example where only the first verb has braced negation and the [neither] and >norC labels are therefore divided between the two negators’ tags:

$/neg-v([neither])>=_NE
$harm/vps13K2_HEARM+Ed $/vps13[N]K2_+Ed
$/P13NI_HIT
$/P12Od-ad_TE
$not/neg-v<=(>norC)_NAw>IH>T
{.}
$nor/cj_NE
$sylian/vps13K2_SUL+Ed $/vps13[L]K2_+Ed
$/P12GOd+C_yI
$soul/nOd_SAwLE

‘It does not harm thee, nor defile thy soul’ (bod34t.tag)

Note that unbracketed words in the negation tags, always with directional arrows, point to eME equivalents that are actually present in the text. The bracketing of [nor] and [neither] in the tags indicates that the bracketed words are implied but not present. Where words in the negative tags are actually missing because of damage to the MS or probable error the relevant labels to the negative words are placed between [[ in the same way as the missing words themselves are identified as being supplied, as an aid to interpretation only, rather than being given as reconstructed Middle English forms:

$/DatN_y~
$bitter/ajn_BITTERE
$last/vps13_I+LaeST+Ey $ge-/xp-vps13_I+ $/vps13_+Ey
$ever/av_aeFFRE
{.}
$/DatN_y~
$sweet/ajn_SwETE
$/neg-v(>[av[)_NE
$come/vps13_CUM+Ey $/vps13[N]_+Ey
$/P12Oi_yE
!_[N][]
{[never[}

‘That bitterness lasts forever. That sweetness [never[ comes to thee’ (worcthfragst.tag)


4.4.6 Adverbs
Tagging is categorical; natural language tends to be fuzzy. Therefore it is impossible to build clines into a tagging system such as ours. Unfortunately, the fuzziness of natural language may affect the assignment of word forms to grammatical categories; in no place is this conflict more difficult than in the classification and analysis of adverbs. The basic label for adverbs is av. Particular words marked av in the tagged texts might well be considered by some to be conjunctions rather than adverbs. Many words, e.g. ‘where’, ‘when’, ‘as’ may function as both adverbs and conjunctions. We label such words with av or cj according to our interpretation of context. The label cj (see further §4.4.8 below) is reserved for simpler examples of ‘classical’ subordinating and coordinating conjunctions.

4.4.6.1 Simple adverbs
The clearest cases of the category adverb are the familiar adverbs of manner, time and place:57

$/P11N_IC
$will/vpt11_wOLDE
$/TOd_dE
$wretched/ajOd_wRECCHEDE
$soul/nOd_SAULE
$so/av_SA
$rueli:ce/av_REw+LICHE $-ly/xs-av_+LICHE
$acwellan/vi_A+CwELL+AN $a-/xp-v_A+ $/vi_+AN

‘I would so ruefully kill the wretched soul’ (vvat.tag)

$&/cj_&
$often/av_OFTE
$do/vps13_DO+d $/vps13[V]_+d
$ofthink/vi_OF+yENCH+En $of-/xp-v_OF+ $/vi_+En
$that/cj_y~
$/P13NM_HE
$ever/av_aeURE
$so/av_SwO
$have/vps13_HAU+ED $/vps13_+ED
$do/vSpp_I+DO+N $ge-/xp-vpp_I+ $/vSpp[V]_+N

‘...and often doth regret that he ever so hath done’ (vvat.tag)

Adverbs may also be given semantic specifiers, especially those that may function also as prepositions:

$fela/ajpl_FELE
$unwine{fr}/npl_UN+wIN+ES $un-/xp-n_UN+ $/pln_+ES
$before{p}/av_TEFOREn
$&/cj_&
$baeftan/av_BAFTEN

‘Many enemies before and behind’ (vvat.tag).

4.4.6.2 Comparison
As with adjectives, comparison is indicated by the labels -cpv (comparative) and -sup (superlative):

$/P11N_IC
$will/vpt11_wOLDE
$blithely/av-cpv_BLIdE+LIC+OR $-ly/xs-av-cpv_+LIC+OR $-er/xs-cpv_+OR
$Yolian/viK2_yOL+IgEN $/viK2_+IgEN
$/TG_dAS
$li:chama/nG_LICHAM+ES $/Gn_+ES
$death/nOd_DEAd

‘I would more happily suffer the death of the body’ (vvat.tag).

$/TplN_yA
$2/qc_TwA
$sin/npl_SUNN+E $/pln_+E
$/RTIplOd_yE
$man/npl_MEN
$follow/vps23_FULI+Ed $/vps23[V]_+Ed
$swi:De/av-sup_ALRA-SwId\+EST $alder-/xp-av_ALRA- $-est/xs-sup_+EST

‘The two sins that men follow most strongly of all’ (lamhomA1t.tag).

4.4.6.3 Coordinating function
The adverbs found most frequently in coordinating constructions are $as and $so. As with $in/pr and $on/pr (see §4.4.7.2 below) these words are not historically distinct. They represent disassemblies of OE eall-swā. We have therefore assign forms to the lexels $as and $so, not on an orthographic but on a semantic basis:

$as/av>=_*AL-SwO
$deeply/av_DIEP+LICHE $-ly/xs-av_+LICHE
$have/vps13_HAF+d $/vps13[F]_+d
$god/n_GODD
$/P21Oi_US
$forbid/vSpp_FOR+BOD+EN $for-/xp-v_FOR+ $/vSpp_+EN
$all/ajplOd_ALL+E $/plajOd_+E
$headsin/nplOd_HEAUED-SENN+ES $sin/nplOd-k_-SENN+ES $/plnOd_+ES
$as/cj<=_SwO
$/P13NM_HE
$do/vpt13_DEDE
'_ADAME
$/TOd_dE
$tree/nOd_TREU
$of/pr_OF
$paradise/n<pr_PARADISE

‘God has forbidden us all capital sins as profoundly as he did Adam the tree of paradise’ (vvat.tag).

In the construction above, one aim of the tagging is to isolate the coordinating pair ‘as..as’ for the purposes of comparison and mapping. The first of the pair therefore has >= added to its grammel and the second has <= (cf. the discontinuous compound relatives described in §4.4.2.3 above and braced negation in §4.4.5.1 above). Here this essentially indicates a braced construction as well as providing flags for a program that links the two in the tag dictionaries. The second element of such braced constructions is normally given the label cj, indicating its conjoining function.58 The choice of av for the first element reflects its frequent simplex usage. Neither label, however, can be taken as categorically precise as the same labels might if attached to items like ‘deeply’ and ‘or’ respectively.
It will be apparent from the example of ‘as..as’ above that the category av is employed in the tagging process as something of a catchall. Other examples of how this label has been used to identify fuzzy-edged notions are in the braced construction ‘both..and’, which is labelled $both/av>= $and/cj<= . The category ‘adverb’, even in the standard grammatical literature, is not as clear or univocal as e.g. ‘noun’ or ‘verb’. Because of its complexity, ‘adverbs’ represent a particularly rich field for realising the protean capabilities of our tagging system. Users are free to alter or elaborate the labels here assigned according to their own analytical preferences.

4.4.6.4 Compounds
Compound adverbs are of two types: (1) adverb + adverb and (2) adverb + preposition. Both elements of type (1) have the grammel av-k. The second element of type (2) has the grammel pr-k, see the examples below and cf. §4.4.7.3 below:

$/DatN_*yAT
$holy/aj_HOLY
$tree/n_TRE[O]
$be/vpt13_WAS
$fair/aj-sup_FAIR+OST $-est/xs-sup_+OST
$Ya:/av_yO
{.}
$/RTIOd_y^T
$/P23N_HI
$may/vpt23_MIzTE
$anywhere/av-k_A+WER $where/av-k_+WER
$see/vi{rh}_I+SE[O] $ge-/xp-vi_I+ $/vi[V]{rh}_0

‘That holy tree was the fairest then that they might anywhere see’ (corp145selt.tag).

$for/cj_FOR
$many/ajOd_MONI
$evil/nOd_UUEL
$/P11N+V_ICH
$see/vps11_I+SEO $ge-/xp-vps11_I+
$therein{p}/av-k_yR+IN $in{p}/pr-k_+IN

‘For many [an] evil I see therein’ (corpart.tag).

4.4.7 Prepositions
The basic label for a preposition is pr. While Middle English is largely a prepositional language, traces of old postpositions remain, especially in verse. In order to facilitate retrieval, and for purposes of comparison, postpositions, which are typically the same lexical items as prepositions, are also marked pr but with an added < .
Since many prepositions in Middle English are highly polysemous, the prepositional lexels will often carry specifiers separating different meanings. Some prepositions will be unspecified, others will have unspecified defaults, while still others will have specifiers in all cases. The assignment of specifiers to prepositions is largely a function of their semantic complexity. Not all polysemous prepositions have been assigned specifiers, (e.g $at/pr, $to/, $for); however, where specifiers are present we have attempted to assign them systematically and exhaustively.59
We will illustrate our use of specifiers with instances of $about:

$/TN_yE
$lion/n_LYOUN
$bray/vpspF-aj_BRAY+INDE $/vpspF-aj_+INDE
$go/vps13_GE+y $/vps13[V]_+y
$about{p}/pr_ABOUTE
$/Dospn<pr>=_yAN
{.'}
$/RTAplOd<=_yET
$/P13NM_HE
$will/vps13_WY\LE
$forswallow/vi_UOR-ZUELz+E $for-/xp-v_UOR- $/vi_+E

‘The lion roaring goes about those that he will swallow’ (ayenbitet.tag).

The specifier {p} means ‘place’ and refers to spatial usages for $about. The next example shows the specifier {t} indicating temporal sense:

$about{t}/pr_ABUTEN
$such/aj<pr>=_SwIC
{\}
$time/n<pr_TIME
$as/cj<=_ASE
$man/indef_ME
$sing/vps13_SING+Ed $/vps13[K]_+Ed
$mass/nOd_MASSE

‘About such time as one sings mass’ (cleoarat.tag),

The sense ‘approximately’ for $about seems to be rare in early Middle English but occurs (labelled {ap}) in Havelok:

$&/cj_AN
$fall/vSpt13_FEL
$/P13NI_IT
$so/av_SO
$that/cj_yAT
$youngman/npl{rh}_YUNGE+MEN $man/npl-k{rh}_+MEN
{\}
$well{w}/av_*WEL
$about{ap}/pr_ABOUTEn
$9/qc_NINE
$or/cj_OR
$10/qc{rh}_TEN
{\}
$beginnan/vSpt23_*BI+GUnN+En $be-/xp-v_*BI+ $/vSpt23_+En
$there/av_yE
{=sic=}
$forto/im+C_FOR-TO
$leika/vi-m{rh}_LAYK+E $/vi-m{rh}_+E

‘And it so befell that young men, about nine or ten, began there to play’ (havelokt.tag).

The sense ‘concerning’ is indicated by the specifier {re}:

$/TN_yAT
$ae:ht/n_AISTE
$&/cj_&
$/TN_yAT
$chattel/n_CATEL
$/RTI>pr_yAT
$/P21N_WE
$about{re}/pr<_ABUTEN
$care/vps21K2{rh}_CAR+ET $/vps21K2{rh}_+ET

‘The goods and the chattels that we care about’ (tr323at.tag).

Note that here ‘about’ governs the relative pronoun yAT and is postposed. yAT therefore is given the grammel RTI with the annotation >pr and $about{re} is given the corresponding grammel pr<.

4.4.7.1 Influence of following words
The form of some prepositions is potentially affected by the initial of the following word. As with the personal pronouns ‘my/mine’ and ‘thy/thine’ (§4.4.2.1 above) the grammel for these prepositions carries +H or +V in these cases. Note that those followed by a consonant-initial word are here unmarked. For instance final <m> in ‘from’ is more common before vowel or <h>, in ‘(in)to’ and ‘unto’ the alternant ‘til’ is sometimes similarly triggered.
A special case is $at when followed by the definite article. Because of the juxtaposition of the two dentals, in many text languages the definite article assimilates to the preceding preposition. In these cases, ‘at the’ is therefore tagged as a single unit. The tag takes the form $at/pr+T:60

$/TOd_*yE
$angel/nOd_ANGEL
$/P12N_yOU
$shall/vps12_SSELT
$there/av_yARe
$find/vi_FIND+E $/vi_+E
{.}
$/RTA_y^T
$drive/vSpt13_DROF
$/P11Od_ME
$out/av_OUT
$at/pr+T_ATTE
$gate/n<pr{rh}_zETE

‘There thou shalt find the angel that drove me out at the gate’ (corp145selt.tag).

Following initial <h> and vowel are indicated in the format bH and bV (b = before):61

$at/pr+TbH_ATTE
$high/aj<pr_HEIE
$noon/n<pr_NON

‘At the high noon’ (corp145selt.tag),

$at/pr+TbV_ATTEN+
$end/n<pr{rh}_+ENDE

‘At the end’ (corp145selt.tag).

4.4.7.2 $in/pr and $on/pr
Old and Middle English did not make the hard and fast distinction between the (historically related) word shapes in and on that Modern English does. In addition, there are reduced spellings, not only <i>, <o>, but also <a>, which could reflect either word. We have therefore decided to assign forms to the lexels $in and $on, not primarily on the basis of spelling, but using semantic criteria. For instance, any spelling, whether <i(n)>-type, <o(n)>-type or <a(n)>-type, that refers to ‘earth’ as a place of habitation is given the lexel $on. When the preposition governs ‘earth’ as a place of burial it is given the lexel $in:

$for/cj_UOR
$/P12N_yU
{\}
$shall/vps12_SSELT
$live/vi_LIBB+E $/vi_+E
$/D-cpv_yE
$long/av-cpv_LENG+ER $-er/xs-cpv_+ER
$on{p}/pr_INE
$earth/n<pr_YERyE

‘For thou shalt live the longer on earth’ (ayenbitet.tag).

$/P13NM_HE
{\}
$a:gan/vps13_AHTE
$to/im+C_TO
$niman/vi-m_NIM+ENE $/vi-m_+ENE
$much/ajOd_MUCHELE
$ge:me/nOd_gEME
$that/cj_y~
$/P13NM_HE
$/P13OdI_HIT
{\}
$sow/vsjps13_SAwE
$in{p}/pr_ON
$becomely/aj<pr_BI+CUME+LICHE $be-/xp-aj_BI+ $-ly/xs-aj<pr_+LICHE
$earth/n<pr_EORdE

‘He ought to take great care that he sows it in proper earth’ (lamhomA1t.tag).

Notice in the examples above the specifier {p} = place. Because of the wide semantic range of both $in and $on (spatial, temporal, illative, manner adverbial...) it is necessary to add further specifiers. For complete lists see the Tag Key.

4.4.7.3 Complex prepositions
Not all prepositions consist only of one word. For instance, the common collocation ‘out of’ is treated as a simplex and given the lexel $outof/pr. Where compound adverbs are made up of adverb + preposition, e.g. $therein, $wherefrom, for purposes of retrieval and comparison, the second element is given the grammel pr-k (cf. §4.4.6.4 above):

$/P13NF_HA
$bind/vps13-ct_BINT
$/P13OdXF_HIRE
$thereto/av-k_yER+TO $to/pr-k_+TO

‘She binds herself thereto’ (corpart.tag).

When compound adverbs made up of adverb + preposition are disjoined, their relationship is indicated by arrows:

$if/cj_zEF
{\}
$there/av>_yER
$be/vps13_IS
$overhogu/n_OUer+HOHE $over-/xp-n_OUer+
{.}
$or/cj_OdER
$any/aj_EI
$pride/n_PRUDE
$in{p}/pr<_IN

‘If there is contempt or any pride in’ (bod34t.tag).

Sometimes the first part of the compound is not expressed:

$/P12N_yU
$open/vps12K2_OPPN+ES $/vps12[N]K2_+ES
$/P11Oi_ME
$/P12GOd+H_yIN
$heart/nOd_HERTE
{\}
$forto/im+C_FOR-TO
$know/vi-m_CNAw+E $/vi-m_+E
$vitrli:ce/av_wITer+LICHE $-ly/xs-av_+LICHE
$&/cj_&
$in{p}/pr<_In+
$to/im+C_+TO
$read/vi-m_RE\D+EN $/vi-m_+EN
$true/ajplOd_TREw+E $/plajOd_+E
$loveletter/nplOd_LUUE-LETTR+ES $letter/nplOd-k_-LETTR+ES $/plnOd_+ES


‘Thou openest to me thy heart to know [it] surely and [there]in to read true love-letters’ (tituswoht.tag).

Here the backward arrow in $in{p}/pr< points back to an empty category.


4.4.8 Conjunctions

4.4.8.1 Simplex conjunctions
The basic label for conjunctions is cj. A conjunction is a particle, which joins two syntactic constituents of similar or identical structure. In the standard literature, subordinating and coordinating conjunctions are distinguished. Our labels do not make the distinction. It must be remembered that grammels are not primarily designed as parsing tools, but are part-of-speech labels enabling comparison of formal differences that may carry temporal or spatial implications. More sophisticated syntactic analysis is nevertheless facilitated by them. Our basic tags may be used as initial pointers to identify subordinate or coordinate constructions and thereafter they may be elaborated or themselves further parsed.
The following example illustrates the subordinating conjunction ‘that’ and the coordinating conjunction ‘and’ both labelled cj:

$say/v-imp22_SEGG+ED $/v-imp22_+ED
{\}
$that/cj_yET
$/TN_yE
$lord/n_LAUERD
$have/vps13_HAU+ED $/vps13_+ED
$thereof/av-k_yAR-OF $of/pr-k_-OF
$need/nOd_NEODE
$&/cj_&
$hraedli:ce/av_RED+LICHE $-ly/xs-av_+LICHE
$/P23N_HEO
{\}
$/P22Od_EOU
$lae:tan/vps23_LET+ED $/vps23_+ED
$faran/vi_FER+E $/vi_+E
$theremid/av-k_yER-MID $mid{w}/pr-k_-MID

‘Say that the Lord hath need of it and quickly they will let you go with it’ (lamhomA1t.tag).

Use of the label cj, like the label av (see §4.4.6 above), is subject to contextual constraints. Sometimes it is very difficult to decide whether a given item should be interpreted as a conjunction or as an adverb. We use substitution to help us decide. If unequivocal conjunctions such as ‘because’, ‘although’ or ‘if’ can be substituted for a given item and make at least grammatical sense, then the item is labelled cj. If unequivocal adverbs, such as ‘like’ or ‘how’ make grammatical sense when substituted, then the item is labelled av. Compare the following two examples:

$sothat/cj_ZUO-yET
$/P13NM_HE
$witannot/vps13_N+OT $n-/xp-neg_N+
{,}
$where/av_HUER
$/P13NM_HE
$be/vps13_YS

‘So that he knows not where (cf. ‘how’) he is’ (ayenbitet.tag).

$nor/cj_NE
$seek/vi_ZECH+E $/vi_+E
$kindly/ajOd_KENDE+LICH $-ly/xs-ajOd_+LICH
$skill/nOd_SCELE
{.}
$where/cj_HUER
$/P13NM_HE
$none/pnOd_NON
$/neg-v<pn_NE
$have/vps13_HE+y $/vps13[V]_+y

‘And not to seek natural skill, where (cf. ‘if’) he has none’ (ayenbitet.tag).

4.4.8.2 Complex conjunctions
Discontinuous constructions such as ‘either .. or’, ‘neither .. nor’ are labelled cj>= .. cj<= to ennable retrieval as single markers (cf. LALME items 111 and 188):

$man/indef_ME
$/neg-v>(neithernor2)_NE
$swear/vsjps13K2[1]_ZUERIE
{,}
$neither/cj>=_NE
$by/pr_BY
$/T<pr_yE
$heaven/n<pr_HEUENE
{,}
$nor/cj<=_NE
$by/pr_BY
$/T<pr_yE
$earth/n<pr_ERyE
{,}
$nor/cj_NE
$by/pr_BY
$other/aj<pr_OyRE
$sceaft/n<pr_SSEPyE

‘One swears neither by the heaven nor by the earth nor by other created things’ (ayenbitet.tag).

Notice from the above example that when the second element of a negative coordinating construction is repeated, only the first instance carries the linking grammel specifier <=. Information about the compound negation is, however, to be found in the tag to the initial negative particle NE, here marked $/neg-v>(neithernor2) (see further §4.4.5 above).
Conjunctions may coordinate not only with other conjunctions but also with adverbs. Labelling these elements as cj or av is again context dependent and therefore in some cases arguable (cf. §4.4.6.3 above).


4.4.9 Interjections
The label for interjections is int. Syntactically they are simple; the only distinction worth noting is that between vocative and non-vocative interjections exemplified below:

$&/cj_*AND
$say/vpt13_SEI+DE $/vpt13[V]_+DE
$alas/int_ALAS
$&/cj_AND
$make/vpt13_MAU+DE $/vpt13[V]_+DE
$cry/nOd{rh}_CRI
{\}
$ah/int_A
$lord/n-voc_LOUerD
$/P13NM_HE
$say/vpt13_SEI+DE $/vpt13[V]_+DE
$scendan/vpp_I+SCHEND $ge-/xp-vpp_I+
$/P11N+V_ICH
$be/vps11{rh}_AM

‘And said “alas” and made cry: “Ah, lord”, he said “I am undone”’ (laud108at.tag).


4.5 Tagged texts

We have described above the theory and conventions for tagging texts, and have illustrated the tagging for particular items. This chapter has been concerned with what the human tagger does; that is the decisions made by the scholar interacting with the textual material. We do not here give documentation for the program ‘Tagger’ that handles the interactive process of tagging, e.g. the hiving off of selected affixes, because we assume the LAEME corpus, however incomplete, is a given. Interactive tagging has been described in earlier publications (Williamson 1992/3, Laing 1994). We hope in the future to make a tagging program available so that users can do their own tagging, whether to add material to the present corpora or to adapt the program for use in compiling their own corpora. The processes of transcription and tagging described in Chapter 3 and in this chapter create the corpus of individual tagged texts.



1 For examples of higher order tagging see Meurman-Solin (2004).

2 Lexels, however, may sometimes be theoretically heterogeneous. An example would be $in{t}/ ‘temporal’, $in{p}/ ‘place’, where the braced label is, because of its universalist reference, distinctly grammel-like.

3 LAEME formatting conventions reserve lower case letters for ‘special’ characters. Here lower case <w> stands for wynn (ƿ), <y> for thorn (þ), <d> for edh (ð), <yx> for barred thorn () = ‘that’.

4 In LAEME format, length, regardless of standard representational conventions, is always marked by a following colon.

5 We assume by default that all French loanwords in our corpus will have been borrowed through the intermediary of some form of French used in England. For arguments justifying such a position see Rothwell (1998).

6 The system for indicating Old English characters in the lexels is the opposite of that used in the cited forms i.e. upper case D, Y, indicate edh, thorn where lower case indicates these letters in the citations.

7 Note that the reflexes of OE ge-, where they survive in verbs, are further specified for context, e.g. as to whether the prefix survives in past participle, infinitive, present tense, etc. The survival of reflexes of OE ge- is of potential semantic interest, as well as being a function of history and regional provenance. By early Middle English the prefix ge- in verbs appears to carry no semantic content differentiating the meaning of verbs with the prefix from those without. For this reason, the lexels for verbs that appear with surviving variants of the prefix are the same as those without.

8 Unfamiliar grammels in this and all following examples will be described in subsequent sections. The reference in parentheses following the translation in each case is to the filename of the tagged text from which the example is taken.

9 For an excellent discussion of the problems involved in identifying nominal compounds see Giegerich (2004).

10 The first element of a typical nominal compound is not separable on the same terms as the second element. The second element is sensibly comparable with simplex forms because in virtually all cases it is the head. The first elements can also be recovered if desired because they will always be the first element of a lexel.

11 This example is a verse line — {rh} attached to the end of the tag indicates occurrence in rhyme position.

12 This follows a practice established in LALME.

13 If animals, e.g. ‘the fox and the wolf’, are characters in fables, natural gender is allowed to override non-humanness. Similarly where ‘objects’ in textual context are personifications in an allegory or extended metaphor, e.g. virtues, or ‘the soul’, they may be treated as gendered. However, personification in such texts may be episodic — the virtue or personified attribute may seem to function merely as an abstraction for part of the text and as a true personification elsewhere. Therefore decisions as to whether pronouns should be tagged with M/F (masculine/feminine) or I (inanimate) may sometimes appear arguable or inconsistent. Early Middle English texts show a cline of usage from survival of a degree of historical gender to complete natural gender. Where decisions about the marking of gender are equivocal it seems to us better to mark iffy cases as I (inanimate) so that the user is alerted to the possibility of retained grammatical gender. The user is always free to retag if he considers e.g. ‘the soul’ to be feminine or ‘the sun’ to be masculine in any particular context. Indeed, it is arguable that personifications should be considered as a separately taggable category.

14 Note that preceding dental stops and, in some scribal systems, preceding <s> can have this effect on other grammatical words beginning with dental fricatives also: e.g. ‘thy’, ‘they’ ‘though, ‘then’. In all such cases the grammel of the affected word will be annotated with -ad or -as. Unaffected words are not given the annotation.

15 Note in this example how the dental stop at the end of the word ‘and’ affects the form of the following definite article even though the word ‘and’ itself is represented by the Tironian sign and is not written out. This suggests that there was an aural element in copying, whether it were from self-dictation or dictation by another.

16 In the text dictionaries such linked grammels are combined giving the tag and form, e.g. $/Datpn..RTA_SE..yx. On the treatment of coordinating structures see further §§4.4.51 and 4.4.6.3 below.

17 Cf also words such as $any, $both, $each, $either, $neither, $none, $some, $what, $which, $who, etc. which are given a lexel and, where appropriate, the grammel /pn. These are mostly words which are interpreted as pronominal in OED in at least some functions and which are semantically richer than personal pronouns or determiners.

18 In the Ormulum, the dative case is still recognisable by virtue of the extraordinary regularity of the metre. There would be an argument for tagging the dative in this one text, but we would have nothing to compare it with for dialectal purposes if we did so.

19 In the following discussion we take the weak verb as the default case, and exemplify most verbal categories from this group. Strong, preterite present and anomalous verbs will be discussed separately in §4.4.4.8 . Our discussion there will focus here primarily on the differences between these and the weak verbs.

20 As with personal pronouns (cf. §4.4.2.1), ‘royal we’ and ‘polite plural you’ are marked by 0 — vps01 and vps02 respectively.

21 [x] is a cover term for both [x] and [ɣ]. In phonemic terms, [ɣ] is an intervocalic allophone of /x/.

22 Note here that we assign the [F] marking according to spelling, not presumed phonology. We wish to mark the use of <f>, because during the early Middle English period its use intervocalically is beginning to wane in favour of <v, u>. The inference is that stem-final <f> implies voicelessness, which might trigger syncope. However, it is impossible to tell whether graphic ‘voicelessness’ predicts syncope or vice versa. The source of this problem is the relatively recent phonemicisation of the fricative voice contrast in the early Middle English period, and the fact that there is sufficient scribal memory of OE <f>-spellings to allow them still to be used in some text languages. In the unsyncopated form, therefore, <f> may well represent intervocalic [v]. Conversely while <u> in BI+LEU+Ed may be taken to represent [v], in BE-LEU+y it is possible that it represents [f] as the ‘underlying’ stem final. In these complex circumstances we have decided to use the [F] marker simply to draw attention to retained <f>-spellings without making a potestatic commitment.

23 Note that syncopated spellings with preceding <w, ƿ> are very rare. In almost all cases of syncope the ‘w’ has overtly ‘vocalised’ and is written <u>. In such cases the detached ending is therefore marked with [V] (for vowel) not [W]. Uncertainty as to whether coda vocalisation has happened or been completed in particular cases is of potential historic interest. It is for this reason that we mark ‘w’ stem finals still written <w, ƿ>: these spellings may imply continuing [w].

24 This form is from the Ormulum and the combination <> in Orm’s orthography always implies [j]. The detached +y therefore has the tag $/vps13[J]. In some text languages the present singular indicative forms for ‘say’ retain <-gg->, e.g. SEGG+Ed; the detached endings in these cases are left unmarked since [dZ] does not normally trigger syncope. In others, the forms show vocalisation with or without a retained affixal vowel (SEI+d/SEI+Ey); the detached endings being marked [V].

25 Note that hortative subjunctives immediately following the pronoun ‘we’ (i.e. the type translated ‘let us’ in Modern English) also appear to have reduced endings (see §4.4.4.6 below).

26 Towards the end of the early Middle English period we begin to find zero endings also in consonant stems but these are very few. If they should be of interest to any user, they can of course be tagged.

27 There is a small class of OE class I r-final weak verbs with a thematic -i-, like that of class II (e.g. herian, derian). To capture both their behaviour and origin we mark them with the complex specifier K2[1]. We mark similarly the weak present forms of the otherwise strong verb swerian. Note also that some verbs marked K2 may originally have belonged to class III weak type (see Campbell 1959: §763); the presence of thematic -i- is what is primarily being noted.

28 Note that in marking French verbal loans in this way, we are not suggesting that the categories are stable in Anglo-French or indeed in Continental forms of Old French. The range of forms cited in The Anglo-Norman Dictionary (even taken from the largely normalised sources used in its compilation) indicate considerable leakage from one conjugation to another. My purpose in marking the grammels (as with the marking of original Weak Class II verbs) is to flag the possible survival of indications of the more archaic -i- conjugations.

29 The vowellessness of the suffix in ‘have’ and ‘live’ could also have a historical explanation. These two verbs are members of OE weak class III, which typically lacks a vowel in this context. However the LAEME corpus suggests that these may not be straightforward survivals. In ‘live’ the -ED(E)-type past tense endings are the norm and in ‘have’ there are two apparent pathways: (a) assimilation of OE -fd- to -dd- giving a past tense that behaves like that of a vocalic stem verb; (b) extension, like ‘live’ to -ED(E)-type suffixes. It is possible therefore that the +D spellings in our examples are reinventions rather than survivals.

30 These endings therefore look the same as those for strong past plurals which are differentiated from them by the addition of S between the v and pt label (§4.4.4.9.1).

31 For discussion of past tense indicative and past participle in strong verbs see §§4.4.4.9.1 and 4.4.4.9.2 below.

32 Note that in the example above, the 2nd person plural pronoun, yE, is taken as going with A+WARI+EDE — ‘ye cursed ones’ rather than with the verb — ‘go ye, cursed ones’. It is therefore left unmarked.

33 The same decision has been made for the past participle endings of these words: see further under (d) below.

34 This would give the combined tag $for..to/im+C.

35 Note that superscript <e> or <i> indicating <re> or <ri> can straddle the stem and the suffix in other parts of the verb, where the stem final consonant does not normally affect the ending, and where the [R] label is not therefore required. In these cases, and also when the phenomenon occurs in the genitive and plural of nouns and adjectives with stem final <r>, the [r] label alone is used.

36 Syncopation even in the past participle is observable more in texts from later in the early Middle English period and in those texts from the North or North Midlands than in those from further south.

37 Cf. also the negative versions of two of them $witannot and $a:gannot for which see further §4.4.5 below. Note also that when $shall and $may function not as auxiliaries but have the full force of a main verb with the repective senses ‘owe, be appropriate to’ and ‘have power, avail’, the lexel is annotated with {mv}.

38 As for $shall and $may above, when $will and $willnot function as main verbs with the sense ‘desire (not)’ the lexel is annotated with {mv}.

39 The lexel $go includes forms in GANG-, GONG-, zONG-. The past tense indicative is however given the separate lexel $e:ode.

40 The verb ‘be’ is actually three verbs historically: an s-stem (which includes eom, etc.) a b-stem, and a w-stem. The lexel $be includes forms from all three stems.

41 For more detailed exposition of the complexities of early Middle English verbal negation see Laing (2002) and references cited there.

42 We do not attempt to assign special tags to negative words that negate clauses without the help of the markers ‘ne’ and ‘not’.

43 In the text dictionaries and on the maps these two tags together yield $ne..not/neg-v.

44 In the text dictionaries and on the maps these two tags together yield $n-..not/neg-v.

45 Unequivocal examples of this construction are few and far between. We discount all ambiguous instances where the first ‘ne’ could be read as a clausal conjunction ‘nor’ (see below §4.4.5.6), even where the ‘ne’ is linking a positive clause with a following negative one and could therefore formally be analysed instead as a ‘ne + ne’ construction.

46 Both these constructions are found in Old English  and continue into early Middle English (Mitchell 1985: §§1604-05, 1608).

47 The two tags together yield the coordinated tag $ne..ne/neg>v.

48 The two tags together yield the coordinated tag $ne..n-/neg>v.

49 The two tags together yield the coordinated tag $not..ne/neg>v.

50 The two tags together yield the coordinated tag $not..n-/neg>v.

51 These tags together yield the coordinated tags $not..ne/neg>v and $not..n-/neg>v. See below, §4.4.5.5 for further discussion of ‘naught’ and its treatment in the LAEME corpus.

52 See further Laing (2002: 312).

53 Ayenbite of Inwyt fol. 91r (Morris 1866: 253 lines 14–16).

54 The combination needs to yield the coordinated tag $ne..ne..not/neg-v.

55 For fuller discussion and references, see Laing (2002: 306–309).

56 Pace Smithers (1987: 103) in his note to line 420 of Havelok, the practice of placing a single negative before the second of two words to be negated may be ‘normal ME usage’ with words other than verbs, and with non-finite categories of verb, but it appears to be extremely unusual in linking two finite verbs to be negated.

57 One adverbial type not treated in this section is the relative adverb. These are construed not as a class of adverb but as a class of relative marker — see §

58 Note however in ‘when.. then’ constructions the labels are usually reversed, reflecting the strong conjoining force of ‘when’ and the temporal adverbial function of ‘then’.

59 In the more polysemous prepositions, there is, however, a great deal of overlap and ambiguity. Some specifier choices will therefore be debatable.

60 Note that only those forms that show true assimilation are tagged in this way. Any form of the definite article after ‘at’ that retains its initial fricative, whether or not there is a space between the two words in the manuscript, is tagged separately. If the two words have been run together, e.g. ATyE, they are tagged $at/pr_AT+ and $/T<pr_+yE where the plus signs show the joining of the words but the separate tags indicate the lack of assimiliation. Even if the initial fricative has become a dental, but the two words are kept separate in the manuscript, this is taken to imply lack of assimilation and is tagged merely to indicate influence of the preceding dental as described in §4.4.2.2 above: e.g. $at/pr_AT and $/T<pr-ad_TE.

61 In early Middle English, the definite article in oblique functions (especially post-preposition) does sometimes retain final <n> (or <r> in feminines) and these survive more frequently when the article is followed by a word beginning with <h> or a vowel. Unfortunately, we have not annotated the tags of unassimilated forms of the definite article when they occur before <h> and before a vowel. It would have been a very good idea, and operationally simple, but having not made the decision to do so at the outset, it would have proved too time-consuming to go back and do so at a late stage. Such differentiation must be left to a second order of tagging.
References