History of Scots to 1700

9. Style

Structures of sound §§1.1-1.2.3 and §§1.2.5-1.3 – CM
Unstressed -e §1.2.4 AJA ed. CM
Verse §§2-4 - AJA ed. CM
Prose §5 - CM

Weill at a blenk sle poetry nocht tayn is (Douglas Aen., Prologue I: 108)

The reader of OSc must depend heavily either on glossaries or on DOST, and there is a great deal to be said for the habit of using DOST. A good glossary has the obvious advantage of being keyed to forms actually occurring in the texts in question, including inflected forms of words, but it can give only the bare bones of meaning. DOST can help the reader to recover something like the contemporary force of a writer's choice of words.

9.1 Structures of sound[175]

Unlike the modern artist, the medieval or Renaissance practitioner was not concerned to break out of the restrictions of inherited conventions and forms, but to master and develop them. The ethos was one of craft skill, rather than romantic individualism. The medieval poet might play with the boundaries between genres, as Dunbar does, for example in The Tua Mariit Wemen and the Wedo, but the motivation was irony or humour, not the confusion or elimination of boundaries for the future. We can therefore regard choices of form and style as reliable guides to the literary tradition within which the poet wishes the work to be read.

9.1.1 Alliteration

9.1.1.1 Pure stress metre

There were two types of verse structure in Middle Scots. One, syllabic verse, was based on metrical feet, i.e. regular sequences of weak and strong syllables. This is the type of metric still most closely associated with the idea of verse, but it is not native to English. It was borrowed in the Middle English period from Old French (and Medieval Latin). Old English verse was structured on a different principle, pure stress metre with alliteration. This native form is based on a regular number of alliterating strong syllables in each line. Rhythmically, it reads like free verse, but in its strict OE form the half-lines are quite short and utilise a limited number of arrangements of weak and strong syllables.[176]

The tradition of writing strictly structured, unrhymed alliterative verse apparently disappears after the Old English period (giving way to looser forms often combined with rhyme), but reappears in England in the 14th century. Some continuity, even if only through the reading of earlier work, is implied by the Alliterative Revival's use of certain items of native poetic diction that are not otherwise preserved, but Turville-Petre (1977) discounts the view that OE traditions somehow survived orally.

The earliest poems of the Revival are unrhymed, but later the alliterative line is combined with rhyming stanzas. In this form, the Revival is taken up by Scots poets in the next century (the earliest extant example being Holland's Howlat),[177] and it continues into the 16th century in Scots, after it had again died out in England. The Revival was strongly associated with the West Midlands of England, and some dialect forms from this area find their way into the Scots contributions to the genre, e.g. wlonk and blonk.

The basic form of the Germanic tradition of pure stress metre, as found in OE, can be described as follows:

  • there are four strong stresses to the line;
  • the first three of these stresses alliterate with each other;
  • there is a strong caesura (medial pause) dividing the line into two half-lines;

This can be illustrated from The Dream of the Rood (the only substantial text in OE surviving from Scotland), several lines of which are inscribed in runes on the Ruthwell Cross:

 
hweþræ
Yet there
 
þer
the
A
fusæ
faithful
A
fearran
from afar
 
kwomu
came
A
æþþilæ
nobles [came]
 
til
to
A
anum
the one.
 
ic þæt
I that
A
al bih[eald]
all beheld

(Sweet, revised edn., 1978: 103).

9.1.1.2 Rules of Alliteration in Older Scots

The basic rules of alliteration are the same as in OE. All vowels alliterate with each other and with /h/.[178] Consonants and consonant clusters alliterate if they share the same initial consonant, e.g. /kw, k, kr/. However, in the case of s-clusters, the first two consonants must both be the same, thus /sw/ alliterates only with /sw/, /sl/ only with /sl/, and so on, but /st/ alliterates with /str/, as does /sp/ with /spr/.

The Alliterative Revival line is longer than the strict OE line.[179] The first half-line may have three rather than two alliterating syllables. The second half-line is shorter than the first, and more likely than the first to conform to the types recognised for OE. An important rule of OE alliteration no longer applies in the rhymed stanzas favoured by Scottish practitioners: the last stress now enters into the alliteration. As Crépin points out, if this had happened in OE, it would have fundamentally changed the character of the line. The line and the syntax are normally in tension in strict Old English verse, with the main syntactic breaks falling at the caesura rather than the line end. In keeping with this, the unalliterated stress de-emphasises the OE line end, whereas rhyme emphasises it.

9.1.1.3 Formulae and tags

The liberal use of tags and formulae is characteristic of alliterative poetic diction. Some of these go back to Old English. Tags are short phrases like on mold which are used, sometimes in a fairly meaningless way, to add an alliterating stress. Holland's in nest (see nest n. 1) is an ad hoc creation on this model. Formulae provide pairs of alliterating words, which (as in Old English) can supply a half-line. The most common patterns in MSc are:

  • adjective and adjective (e.g. brathly and bricht)
  • adjective to verb (e.g. feye to be fellit)
  • adjective preposition noun (e.g. thraly in thrang)
  • superlative preposition noun (e.g. blythest of ble).

Obviously, phrases of this kind can readily be constructed. They are described as formulae because many are recurrent and are found in more than one alliterative poem. Mackay (1981) found 186 occurrences of such formulae in the Howlat (many repeated).

Conventional formulae are well-known as a feature of oral composition, allowing pre-literate bards to ad lib, at least partly, in the rendering of traditional material. Holland's need for such assistance comes from another source: the peculiarly demanding poetic form he has chosen: the alliterating and rhyming wheel stanza (see below). The usual criticism levelled against such writing is that sense takes second place to the exigencies of composition. Another symptom of this is the stretching of the meanings of individual words:

The style, even for that time, is particularly uncouth, from the constant alliteration and consequent necessity of using old and uncommon words ... The reader will be quite satisfied with the HOULAT as a specimen of this counterfeit language, formed more for the purpose of sound than sense. (Sibbald, 1802, quoted by Mackay, 1981: 191)

In the latter part of the Middle Scots period, pure stress alliterative verse was relegated to comic and vituperative subjects.[180] James VI in The Revlis and Cautelis (Poems I 75) calls it “tumbling verse” and regards it as suitable “for flytings”.

In a way, the strengths of alliteration are also its weaknesses. The sounds it focuses on are repeated over a much shorter space than with rhyme, and are, moreover, usually consonantal, so the sound effects are very marked. This lends itself readily to phonaesthesia (see below), but may be felt as too insistent when no such special effect is aimed at. In the Revival works, the alliteration is often treated as a challenge, and the poet takes pride in the fullness of alliteration and its maintenance over several lines. Alliteration has typically to be supplied in greater quantity than rhyme, putting a severe strain on the poet's vocabulary. This is partly met by resort to poetic diction, but there is also a strong temptation to become repetitive.

In the Revlis, James VI writes, “let all ȝour verse be Literall [alliterative], sa far as may be”, and indeed alliteration is very often used as an ornament even in syllabic verse. Barbour uses it, for instance, to heighten descriptive passages, and Henryson's use of it in some basically syllabic verse works, for instance "The Ressoning betuix Aige and Youth", is so systematic as to be "almost structural" (Riddy, 1988: 42, citing Denton Fox). Kiparsky (1981) suggests that the rhythmic tendency of English inherently favours alliteration. The stress always falls on the root syllable of a word, e.g.:

/
tie, 
 
un
/
tie, 
/
tieing.

This is in contrast to the mobile stress of the Romance languages, seen in loans such as:

/
obligate, 
 
ob
/
ligatory, 
 
obli
/
gation.

In Scots, then (as in English), to emphasise the stressed syllable is generally to emphasise the root of the word, and alliteration achieves this more readily than rhyme, given the trochaic tendency of Scots and English.

9.1.1.4 The wheel stanza

The wheel stanza is a modification, preferred by Scots poets,[181] of the bob-wheel stanza of some Middle English alliterative verse. In the latter, there is a short bob line (of one stress) at 1.9 instead of a long line.[182] The stanza can be divided into two parts: nine long lines, followed by a wheel. The last two lines of the wheel form, in effect, one long line (the first half-line tends to be longer than the second), and sometimes they do also alliterate with each other. This stanza form can be illustrated by stanza five of The Howlat:

 
He
A
grat
A
grysly
A
grym
 
and
A
gaif
 
a
A
gret
 
yowle,
A
Cheverand
 
and
A
chydand
 
with
A
churliche
A
cheire.
 
“Quhy
 
is
 
my
A
fax,”
 
quod
 
the
A
fyle,
A
“fassonit
 
so
A
foule,
 
My
A
forme
 
and
 
my
A
fetherem
    A
unfrely,
 
but
A
feir?
 
My
A
neb
 
is
A
netherit
 
as
 
a
A
nok,
 
I
 
am
 
bot
 
ane
?A
owle!
 
Aganis
A1
natur
 
in
 
the
A1
nicht
 
I
A2
walk
 
into
A2
weir;
 
I
A
dar
 
do
 
nocht
 
on
 
the
A
day,
 
bot
A
droupe
 
as
 
a
A
doule,
 
Nocht
 
for
A1
schame
 
of
 
my
A1
schape
 
in
A2
pert
 
till
  A2
appeir.
 
Thus
 
all
 
thir
A
fowlis
 
for
 
my
A
filth
 
has
 
me
 
at
A
feid
 
That,
 
be
 
I
A
seyne
 
in
 
thar
A
sicht
 
To
A
luke
 
out
 
on
 
day
A
licht,
 
Sum
 
will
 
me
A
dulfully
A
dicht,
 
Sum
A
dyng
 
me
 
to
A
deid.”
 }
 

(from Bawcutt and Riddy eds., 1987: 48)

9.1.2 Metre

Ȝhe writaris all and gentill redaris eyk,
Offendis nocht my volum, I beseik,
Bot redis leill and tak gud tent in tyme.
Ȝhe nother maggill nor mysmetyr my ryme,
Nor alter not my wordis, I ȝou pray. (Douglas, Aen. Epilogue ll. 21-5).

The most popular form of stress-syllable metre in OSc is the iambic pentameter, i.e. a line of five stresses. A wide range of metrical licences were permitted (see below), but even when these are taken into account, there are many works where the five-stress line seems to be varied from time to time with a four-stress one (Bawcutt, 1998: 15), perhaps suggesting a lingering influence of the older metrical tradition (McClure, 2005).

9.1.2.1 The foot

The metrical foot treats syllables as either weak or strong, as if there were only two degrees of stress in spoken language. In practice, at least three levels of stress can be distinguished, and it is often useful to distinguish four, while many linguists would regard stress as a continuum, and argue that there are as many levels of stress as there are syllables uttered in one breath-group. The result is a constant tension between the superimposed pattern of feet and the more various raw material, with subtle effects. Here we consider only the structural rules that govern the rhyming practice of OSc poets.

The most popular type of foot in Scots verse (as in English) is the iamb (x /), followed by the dactyl (x x /). Both combine well with rhyme, because the line ends with a strong beat.

9.1.2.2 Variable and wrenched stress

The preference for iambs (and to a lesser extent dactyls) is a source of tension against the spoken language. Most native disyllabic words are trochaic, e.g. other, bishop, Andrew, thankit, thankand, thanking, thankis. The exceptions are mostly words with prefixes, e.g. amang, before. The trochaic tendency of the language presents difficulties in the supplying of rhyme. (This was not a problem in French, from which this form of verse was borrowed. The loss of many inflections in French meant that the stress came to fall on the final syllable of many words in any case, giving the language an iambic tendency.)

In the modern period, Scots (like English) has tended to solve this problem by tolerating weak or feminine rhyme, e.g. leading : needing, with an extra-metrical weak syllable at the end of the line as required. In earlier times, strong or masculine rhyme is the rule, and this is met by wrenching the stress if necessary, and allowing near-rhymes of short with long vowels. These practices are more evident in a work like The Bruce whose diction is largely native. Consider the following rhymes from The Bruce (references are to McDiarmid and Stevenson's 1980-85 STS edn.):

sturdely: worthi (XVII: 309-10)
thing: defending (XVII: 351-2)
heyr: planer (I: 623-4) (Vowel 2 with Vowel 16)
was: cas (II: 23-4) (Vowel 17 with Vowel 4)
cf. passe: was (I: 629-30) (Vowel 17)
raid: haid (II: 175-6) (Vowel 4 with Vowel 17)
prys: is (XII: 125-6) (Vowel 15 with Vowel 1)

Later, in Middle Scots, rhymes like these are not quite so common with native words, and are perhaps conventional when they are used, but it would seem that Romance suffixes continued to have full vowels.

Such wrenching is reinforced by genuine variation in the stressing of loanwords from French. Most disyllables from this source must have been borrowed as iambs, but this conflicted with the native language, and gradually most of them have been made trochaic. For a long period, certainly in poetic usage, such loans remained variable, so, for instance, a MSc poet could stress natur(e or river (q.v.) on different syllables on different occasions. When Barbour treats AN loans as end-stressed, it is not unlikely that these were living pronunciations, since he is still fairly close in time to the period of borrowing. Cf.:

towne: baroun (II: 81-2)
se: powste (II: 99-100)
wys: seruice (II: 173-4)
auentur: sture (XII: 91-2)
ner: maner (XII: 105-6).

9.1.2.3 -is as an optional syllable

The inflectional ending -is was originally pronounced as a syllable /is/. In Early Scots this was beginning to be reduced to a mere consonant. (But even now in ModSc as in StE, -es is still syllabic after the consonants /s, z/.) Poetic licence allows the poet to follow speech (and reduce the inflection) or tradition (and retain it as a syllable), a convenient way of adjusting the metre (see §6.16). By the second half of the 16th century, however, "copyists were failing to recognise this practice" and were inclined to compensate by adding an extra word (Bawcutt, 1998: I, 15).

The corresponding optional syllable in Chaucerian verse was -e, but this was lost earlier in the North, and does not normally play this role in Scots verse (but see below). A curious exception is the word jugement, which is generally counted as three syllables in Early Scots verse. However, this anomaly was corrected in Middle Scots by changing the word to jugisment.

9.1.2.4 Unstressed -e in 15th-century verse

This sub-section is taken from Aitken (2002).

In certain ESc and early MSc poems in iambic metre there are many instances of the juxtaposition of fully stressed syllables where regular metrical convention requires an intervening unstressed syllable. To remedy this, some editors of these poems have postulated that in such cases an unstressed vowel, usually but not always orthographically present in the text as <-e>, be supplied for pronunciation also. In some instances this is a very attractive supposition, for it regularises what would otherwise be very implausibly irregular lines: e.g. The scharp, grene, suete jenepere (Kingis Quair §32); ? read: The scharp[ë], grenë, suetë jenepere, where ë indicates that an unstressed vowel is inserted in pronunciation.

This option is most regularly resorted to in the strongly English-influenced Kingis Quair, just quoted, most commonly for adjectives, mostly ‘weak’, attributively preceding a noun with an initial stressed syllable, as: The fairest or the freschest ȝong[ë] floure (Kingis Quair §40), but also in other functions, often with no etymological justification: the pryncë than the page (ibid. §9), the rypënesse of resoun lakit I (ibid. §16), bot fourë greis evin (ibid. §21), Nere by the space of ȝeris twisë nyne (ibid. §25). Other poems in which this feature has been plausibly detected are The Quare of Jelusy, Lancelot of the Laik, The Buke of the Chess, and The Buke of the Sevyne Sagis, and there are rare apparent instances in Dunbar, e.g. in The Thrissill and the Rois: With that annone scho send the swyftë ro (l. 78, quoted from Kinsley ed., 1979). But there are no evident grounds for assuming this phenomenon in most other OSc verse, from Barbour onwards, in which, in the main, the circumstances calling for the added /-ə/ are seemingly avoided. We are probably justified in seeing in these belated ‘schwa-supplemented’ forms a Chaucerian anglicised feature of some OSc verse (cf. §9.3.1), perhaps first introduced in The Kingis Quair.

[However, Priscilla Bawcutt (1998: 14-15) is sceptical of such a use of final -e by Dunbar or his contemporaries, and in her 1998 edition of Dunbar does not follow earlier editors in their emendations to introduce ë.]

9.1.2.5 Reduction of –it

The inflection -it can be reduced to a consonant if the root ends in a vowel, e.g. cryit. This was probably already the spoken form in early MSc. The ending is also optionally reduced after nasals and liquids, e.g. answerit. After the fricative consonants, metrical licence already allows the inflection -it to be reduced in the second half of the 15th century (cf. van Buuren, 1982: 112) (but it generally remains syllabic into the modern period after the plosive consonants, e.g. stoppit, biggit).[183]

9.1.2.6 Cuttit short rhymes

The loss of /v/ in words like deil ‘devil’ and our ‘over’ reduced the number of syllables by one. This may be concealed by spelling. Although used with restraint (and apparently regarded as colloquial), such cuttit short word-forms are available for occasional metrical licence in certain genres (see §9.3.7).

9.1.2.7 Other metrical licences and adjustments

Other metrical licences and adjustments are familiar to the modern reader: the substitution of trochees (/ x) for iambs (x /), though not usually more than once in any line (unless for a special effect), and most often in the first foot; elision when two vowels come together; the assignment of syllabic consonants to the preceding syllable, e.g. a word like open treated as a monosyllable ending in a consonant cluster /pn/. The word and, presumably pronounced /n/ in practice, is sometimes treated in the same way. [184]

9.1.3 Phonaesthetics

The more drumly, the more d's; the more stormy, the more s's is almost a formula of Scottish poetry. (Coldwell: 1957: I, 69-70)

Phonaesthesia is a less concrete phenomenon than onomatopoeia, whereby speech sounds imitate other sounds, e.g. the association of /s/ with the hissing of a snake or the whistling of an arrow (see Bawcutt (1976: 157)). It is possible to say that certain sounds have the potential to produce certain effects, but we must beware of drawing up comprehensive schemes and expecting them to hold in every case. For instance, contrast Holland's line:

The bemes blythest of ble fro the son blent (Howlat 3)

with Douglas:

Bank, bra and boddum blanchit wolx and bar (Douglas, Aen. Prologue VII: 57)

both using /b/ and /b/ clusters. Even a sound (or sound-cluster) that seems to be strongly associated with a certain area of meaning is never absolutely tied to a given effect. For instance, the cluster /gr/ occurs in words like growl and groan and has an appropriate force in the line:

He grat grysly grym and gaif a gret yowle (Howlat 53)

but not in:

With gers gaye as the gold and granes of grace (ibid. 28).

Sound symbolism or phonaesthesia is a potential in language that may or may not be activated in particular words and particular contexts, and is readily over-ridden by the (much more salient) senses of actual words.

With these reservations, it can probably be agreed that some consonants (traditionally described as ‘hard’ or ‘harsh’) are noisier, in acoustic phonetic terms, than others. The plosives produced with the tongue, i.e. /t, d, k, g/, and the sounds /ʧ/ and /ʃ/ have this quality. This applies also to clusters including these sounds. Such sounds are favoured in ugsome passages (see §9.2.9), e.g. in stanzas 5 and 6 of the Howlat. The liquids and nasals are less noisy than plosives and fricatives. Consonant clusters are both noisier and more difficult to pronounce than sequences of alternating consonants and vowels. If they are frequent, the effect will be ‘harsher’ and the words will seem to flow less smoothly. This applies also when consonants come together across words, e.g. /tgr/ in grat grysly. Back vowels have a deeper resonance, and these are more often associated with ugsome passages, but in general, vowels are more euphonious than consonants – they lack the bursts of random noise that characterise consonants acoustically.

Particular consonant clusters are vaguely associated with areas of meaning, usually to do with emotions, social disorder, exaggerated physical movements, and other transient and idiosyncratic phenomena. The sound can then be said to symbolise an area of meaning. The teeth-gritting /gr/ cluster was mentioned above as having phonaesthetic force in some words in which it occurs. The /sk/ and /skr/ groups are likewise associated with an area of meaning, which could be described as ‘disagreeable’ or ‘disordered’. Words beginning with /sl/ are often altered to /skl/ with intensifying effect (see SND Introduction §69). Examples, as well as counter-examples, can be found by browsing in the relevant parts of DOST, SND or CSD.

9.2 Older Scots verse styles (Aitken, 1983)

This sub-section (9.2) and the next (9.3) are a summary of Aitken (1983), q.v. for more detailed descriptions of the various categories of verse, and a full account of the membership of each category. Here, only a few well known texts are mentioned as exemplars of each category. Although these sub-sections have been heavily edited, only one substantive change has been made to Aitken (1983): the term Anglo-Scots is reserved here for texts in a thoroughly mixed language such as James I's Kingis Quair, rather than Aitken's wider application to texts here described as anglicised (see §2.5.1). The useful term 'Pre-Standard English' (PreStE), introduced in Aitken (2002), is also employed here.

9.2.1 A theory of Older Scots verse styles

Underpinning the theory of OSc verse styles outlined below is the belief that the MSc poets shared a system of modal decorum to which all of them adhered fairly faithfully. So too did their predecessors in the ESc period, though their system was a simpler and more limited one. On closer examination it turns out that even such an apparently eccentric work as Colkelbie Sow conforms to the MSc system, as Jeffery (1981: 207) has shown.

The following categorisation of OSc verse modes, according to criteria of theme, metre and style, broadly resembles the schemes of Lewis (1954: 68-76) and Ellenberger (1977: 71-5), and the much earlier one implied by George Bannatyne's divisions of his Bannatyne Manuscript collection.

9.2.2 Plain narrative verse

This category includes, for instance, Barbour's Bruce and most of the Wallace. It is at first in tetrameter couplets, and later more usually in heroic couplets. These poems are unelaborate in syntax, and employ plain vernacular language, except for:

  • some of the diction of poetic synonyms shared with alliterative verse, and some favourite formulae and tags used especially in combat episodes;
  • very occasional passages of heightened rhetoric and courtly diction.

9.2.3 Alliterative verse

This category includes the three long narrative poems in the wheel stanza (see §9.1.1.4): The Buke of the Howlat, Golagros and Gawane and Rauf Coilȝear. These again employ a mostly plain vernacular language. The alliterative poems, however, are much more pervasively laced with poetic diction, specifically an alliterative diction of words and formulae that are characteristic of late medieval English as well as Scottish alliterative verse (see Mackay, 1975, 1981). We might also include here Dunbar's The Tua Mariit Wemen and the Wedo (in alliterative blank verse); but on account of its content and other features of style this might rather be assigned to the low-life category.

9.2.4 Middle styles

These next three categories largely eschew the stereotyped diction of the plain narrative and alliterative modes, and draw only incidentally and relatively sparsely on the staple vocabulary of the low-life poetry. The diction and rhetoric of the courtly poetry appear more often, but in much less profusion than in the courtly set pieces, and sometimes seem to be used allusively, as if borrowing or quoting from the courtly mode. The content of the abstractly didactic pieces and passages naturally attracts a high density of latinate vocabulary. Conversely, anglicised forms seem most favoured by the narrative verse. Overall, the style level in these three kinds of verse is middle to high, closer to courtly verse than to other categories.

9.2.4.1 Elaborate narrative verse

This category includes, for instance, the narrative parts of Henryson's Fables, Testament of Cresseid and Orpheus and Douglas's Aeneid (notably wide-ranging and eclectic in language: see Bawcutt, 1976: ch. 6). The members of this category are all in Chaucerian metres, either rhyme royal (see Pearsall, 1962: 58), or heroic couplets, and all are more or less anglicised.[185] The poems in this category are more wide-ranging in lexis and more elaborate in syntax, rhetoric and, in many cases, metre, than the simple narrative verse. They form a half-way house between the latter and the narrative of courtly allegories and visions.

9.2.4.2 Instructive and hortatory verse

This category includes, for instance, a number of early works mostly from Cambridge University Library MS Kk 1.5, No. 6 (the Ratis Raving MS), and also e.g. Henryson's moralitates, and moralising introductions, digressions and culminations in other works.

Similar in tone and content are numerous relatively short pieces, notably the majority of those occupying the first two parts of the Bannatyne MS (fols. 1-96).[186]

9.2.4.3 Reflectively personal poems

A wide variety of stanza forms, some quite complex and many in shorter lines than the Chaucerian iambic pentameters obligatory for elaborate narrative and courtly verse, characterise reflectively personal poems such as Dunbar's ‘Into thir dirk and drublie dayis’, and most of the pieces in the opening section of the fourth part of Bannatyne's collection.

9.2.5 Courtly verse in the grand manner

This category comprises passages and entire poems in what Lewis calls “the full-blown high style” (1954: 74). Poetry in this manner includes several elaborate dream-allegories more or less saturated with classical allusion, e.g. Douglas's The Palice of Honour; and somewhat simpler love-allegories, dream-visions and 'debates', e.g. parts of Dunbar's The Goldyn Targe and The Thrissill and the Rois; and ballats of Our Lady. Almost all of this is in more or less elaborate Chaucerian stanzas. Set pieces in the courtly manner are also prefaced or appended to, or introduced into, works mainly in the narrative and didactic modes, e.g. Henryson's Fables.

This sort of verse draws repeatedly on certain pieces of 'business', including verdant summer-morning descriptiones loci amoeni, which feature a large number of clichés of descriptive detail (many of them virtually invariable or obligatory), embodied in equally recurrent formulae: e.g. the various colourfully gleaming jewels (beryl, topaz, etc.) to which the dew or the flowers or the sunbeams are compared, the tender shoots of the trees in which a bird or birds (Venus' choristers) sing(s) 'from the spleen', the floral 'garth' or garden.

All of this clearly derives, at least in part, from earlier works in some of the same stanzas and a similar manner (albeit with rather less profusion of cliché and lower concentration of formulae) by Chaucer (such as Anelida and Arcite, The Parliament of Fowls and several of the short poems), Lydgate (whose Complaint of the Black Knight is known in the Scottish sources as 'The Maying and Disport of Chaucer'), and Hoccleve (whose Moder of God or 'Oracio Galfridi Chaucer' is also known from Scottish sources). Despite the arguments by Lewis (1954: 74-5) and others against the use of the term, there is a good case for applying the designation 'Scottish Chaucerian' (or, still more aptly, 'Scottish Lydgatian'[187]) to this particular branch of OSc poetry, since almost all of its typical superficial features result from quite conscious imitation by the Scottish poets of the characteristics of these English works, especially those of Lydgate. Though many of these Scots pieces display considerable originality, both in spirit and in technique, this is nevertheless much the most derivative kind of OSc poetry.

9.2.6 Low-life verse

This category, corresponding to Lewis's 'comic poetry' (1954: 69ff.), lies at the opposite pole from courtly verse, and might be described as 'anti-aureate'. It comprises a varied class of burlesque, comic and vituperative poems, a large sample of which is included by Bannatyne under 'mirry balletis' (Bann. MS 98a-211a). These include flytings and lampoons and a number of highly realistic pieces with a low-life setting, including Christis Kirk on the Grene, and the simple rural comedies, The Wyf of Auchtirmwchty and The Wowing of Jok and Jynny.

One must also include here a group of superficially highly unrealistic poems, such as the Ballad of Kynd Kittok. The main point of these poems is to bring the fantastic down to earth by associating it with homely persons, settings and objects; and they are stylistically of a kind with the other poems of this class.

Many of these poems have approximate antecedents in Middle English, such as the northern alliterative Tournament of Tottenham, Chaucer's Sir Thopas, and Chaucer's, Lydgate's and Hoccleve's passages and poems of personal abuse. But in their language and diction (see below), these are the most distinctively Scottish among all the kinds of OSc poetry. Many of the individual poems are also distinctively Scottish in the parochialism or domesticity of their allusions.

Chaucerian stanza-forms are for the most part eschewed in this category.

Some of the typical stylistic features of this group of poems (see below) are shared by many incidental low-life passages, and also horrific passages, within other poems.

9.2.7 Verse of denunciation, protestation and petition

This category includes poems entitled (in original editions and/or by modern editors) 'complaints', 'supplications', remonstrances' and 'petitions'; pieces or passages of denunciation; and many of the poems in the latter half of the fourth part of Bannatyne's collection. Most of these are in tetrameter lines in couplets or simple stanzas, or in other non-Chaucerian metres. They are comparatively simple in syntax, and employ a vernacular diction less densely northern and Scottish than that of the flytings and other personal invectives, but equally low in latinisation and almost devoid of anglicised forms. This style approximates, therefore, to that of low-life verse.

9.2.8 Realistic nature verse

There is a small body of more or less realistic nature verse, namely Henryson's description of the seasons and the countryside in The Preaching of the Swallow and his brief winter setting of The Testament of Cresseid, and the winter scene of Douglas's Prologue to Aeneid VII. These are virtually free of clichés, but later the winter passages too become stereotyped, as in Lyndsay's Prologue to his Dreme, where he laces the winter description with locus amoenus formulae and diction, ostensibly by way of contrast.

The syntax of these pieces resembles that of simple narrative verse, and they employ a predominantly unpretentious vernacular vocabulary adapted to their particular subject-matter.

The narrative modes and the low-life verse also include a considerable amount of more or less realistic dialogue in a level of style not far removed from that of the narrative itself.

9.2.9 Limitations and exceptions

The scheme outlined above will accommodate virtually all surviving OSc verse down to the reign of James VI, allowing for some hesitation between adjacent categories, and some give and take for particular pieces, for instance parodies such as Dunbar's Dregy, and bearing in mind that many poems or passages are not wholly consistent in mode. Nor are the categories watertight; none has an exclusive monopoly of its salient stylistic features (as described below). Descriptions of heaven or paradise naturally attract the language, diction and rhetoric otherwise characteristic of courtly verse, e.g. Adam and Eve in Paradise in Lyndsay's Monarche (785ff.); and conversely, mentions of Hell or of fearsome matter are accompanied by diction and phonaesthetic effects like those of anti-aureate verse. The most densely latinate kinds of verse are the solemn discursive and the courtly; but occasional latinate expressions occur in virtually all other kinds of verse, the low-life narrative pieces only excepted. Equally, although items of heroic (narrative) diction, such as bern n.2, wy and brand n.2, are most often found in plain and alliterative narrative verse, there are nonetheless also stray occurrences of these items in courtly and didactic or mock-courtly works such as The Quare of Jelusy. Some similar strayings out of context of reduced forms (characteristic of low-life verse) are mentioned below.

Style-switching within a work is also normal. This was hardly an invention of the MSc poets. Their English and Scottish predecessors and some English contemporaries knew something of this, especially the northern English. Noisy passages interrupt calmer narrative or dialogue in for instance Sir Gawain and the Green Knight and many of the northern miracle plays and alliterative poems. In Scots, Wyntoun has a number of violent and noisy passages, including a highly alliterative flood (following the northern Middle English Cursor Mundi) at I 397-409. But since they enjoyed an even more extreme range of diction, it is especially in the MSc poets, from Holland on, that the distance between the stylistic poles is greatest and the stylistic contrasts most striking. Furthermore, to these contrasts there is added contrast in onomatopoeic effect between ugsome and amene passages. In Holland's opening six stanzas, the first three stanzas are in amene style, with some native courtly diction and a little latinate vocabulary, displaying complex interlinkings of alliteration on liquid consonants, and rhyming in -ene, -ede, and other 'calm' or 'gentle' phonaesthemes, whereas the second three stanzas on the Howlat's repulsiveness contrast in all these respects (Mackay, 1975: 250-7). Passages of these sorts are among the most heavily alliterated even in verse forms where the alliteration is not structural.

As well as anti-aureate passages embedded in courtly or elaborate narrative poems, there are a few examples also of spoof-courtly passages in comic or low-life settings: for example, the three minions in Lyndsay's Satyre, whose talk is normally in the full colloquial manner, break into courtly diction to describe the attractions of Dame Sensuality (lines 331ff.).

9.3 Types of diction (Aitken, 1983)

9.3.1 Anglicised forms

Anglicised forms are imitated from southern English usage (where they are[188] the regular forms). They are typical features of the poems that Jeffery (1978, 1981) calls Anglo-Scots, namely The Kingis Quair and Colkelbie Sow. They also occur as options in poems of every kind, with the following exceptions:

  • low-life verse,
  • one alliterative poem, Rauf Coilȝear,
  • a number of early couplet narrative poems, including Barbour's Bruce, the Legends of the Saints, Wyntoun, the Asloan Sevyne Sagis and the Prestis of Peblis.

These restrictions in their distribution seem securely to identify anglicised forms as literary, non-Scottish and non-vernacular.

In most of the poems in which they occur at all frequently, most or all of the anglicised forms appear to be authorial. However, it seems that the quite copious (and not wholly coincident) anglicisations of the two texts of the Scottish Troy-book are mostly or entirely post-authorial (McIntosh, 1979, takes a somewhat different view). At least, there is only one anglicised rhyme in the poem that is not also a rhyme when translated into ESc (for example, we still have a rhyme when we replace the text's mo 'more' : two (II: 421-2) with ma : twa): the one exception is slo 'slay' : þo 'then' (II: 2597-8), which looks like a genuine piece of authorial anglicisation. The Troy-book anglicised forms include hyper-anglicisms (see below). We also find the sME hem(e) (for thaim 'them'), otherwise unrecorded in OSc.

In other poems also it is certain that copyists have introduced anglicised forms. But it is also true that in every poem that is at all anglicised, some and often most of the anglicised forms are likely to be authorial. This is evident from the fairly frequent rhymes that will chime only if one of the rhyme-words has the anglicised form, e.g. Scots glore and anglicised sore (e.g. in Dunbar's Ane Ballat of our Lady); Scots donk and anglicised ronk, bonk, thonk (in Dunbar's The Goldyn Targe 93ff.).

The different types of anglicised forms are distributed along an implicational scale. The least anglicised poems of all - the early narratives in octosyllabic couplets - confine their anglicisations to:

  • forms such as one, allone rhyming with personal names such as Jhon, Sampsone, Babilone, and go with Nero, Cupido; indeed, Barbour offers in all one single example of this: Johne : ilkone (XI: 382, in MS C; MS E has Jhane : ilkan). (The form more, confirmed in rhyme in the same works, has a different explanation as a 'genuine' Scots form, see §4.2.1 and relevant dictionary entries).
  • With the exceptions mentioned above, other OSc poems employ a much more extensive range of these 'o for a forms' in Vowel 4 words, e.g. quho for quha, go and gone for ga and gane, cold for cald, etc.
  • Both throughout the anglicised canon and even, for two common words, no and so, in those pieces that otherwise eschew anglicised forms, o for a forms are common within the verse-line; in this case we can of course only surmise whether the choice is the copyist's or his original's.

Many poems of the plain narrative mode (such as the Wallace), alliterative narrative (such as the Howlat), and elaborate narrative (such as most of Henryson), and almost all of the didactic and lyric verse, confine their anglicisation to this feature; and in the more vernacular of these it is apparently merely a rhyming convenience.

Somewhat more anglicised pieces, such as Dunbar's and Lyndsay's courtly poems, add to this several other anglicisation features:

  • the inflection -n on vowel-final stems, in imitation of the PreStE -en ending of the infinitive and the plural present indicative, another valuable rhyme-extending facility: e.g. bene (for be), seyn (for se), sayn (for say v.1); and more rarely fleyn (for fle v.2), leyn (for ley v.), and gane (for ga);
  • rather more rarely, and mostly not in rhyme, the same works have present indicative verbs inflected in -ith (after PreStE -eth in the third person singular present, and, in some dialects, the plural).

In both these instances the Scots poets not infrequently blunder in employing these inflections.

  • a range of periphrastic constructions with the auxiliary do as a metrical convenience (to gain a syllable and/or to shift the main verb into the rhyme position) (see §7.8.15);
  • words borrowed from the English poems of Chaucer, Lydgate and others, and met with in Scots chiefly or only in elaborate narrative and courtly poetry, e.g. frome (for fra, anglicised fro), lyte (for litil(l), morrow or morrowing (for morn(e or morning), tho (for than), and twane (for twa);
  • hyper-anglicisms (forms that do not exist in PreStE, but might have been supposed to, on the analogy of regular correspondences), mostly confined to still more thoroughly anglicised works, such as the Troy-Book and Colkelbie Sow, e.g. o (for a indef. art.) and mod (for made).

These poems display an extensive range of further anglicised forms, including mich (for mekil(l), quich(e) (for quhilk) and sche (for s(c)ho pron.).

The Quare of Jelusy and Lancelot of the Laik add still others, including:

  • individual lexical items such as aȝhane (for again), schall (for sal(l) and shude (for suld (s.v. sal(l));
  • frequent, optional use of the infinitive and present tense inflection -in on consonant-final stems, supplying an additional unstressed syllable, e.g. “To gladin hir and plesyn . . . with thair chere” (Quare of Jelusy 129);
  • uninflected past participles with or without the prefix i-, y- (likewise imitated from PreStE usage), e.g. iclosit (Lancelot 53), y-fret (Quare of Jelusy 548). The only uninflected past participle at all widespread in Scots (not however in the most vernacular texts) is forlore 'forlorn'. Otherwise forms of this sort are confined to this group of ultra-anglicised poems and to Gavin Douglas.

Douglas mostly anglicises much as do Henryson and Dunbar,[189] but he does also make free use of -in verb forms, e.g. “And frely may behaldyn or espy Tha lakis quhilkis thame langis to vissy” (Aeneid VI v 73-4); and occasional use of sche in rhyme (e.g. at Aeneid XI xi 136) and of past participles such as schaw (for schawin) and ytak (for taken or tane) (see further Bawcutt, 1976: 144-5).

On final -e as an optional syllable, see §9.1.2.4.

Without doubt all of these anglicised forms result from an original impetus by Scots poets working in the relevant modes to be at one with, imitate and adapt from, the English masters whom they so admired and imitated in other ways, as well as to benefit from the additional rhyming and metrical resources these practices provided. It should be borne in mind that they were adapting to partially scotticised copies (such as those in MS Arch. Selden. B 24, the Asloan MS and the Chepman and Myllar prints).[190] Nor were they concerned that their imitations should be philologically perfect.

9.3.2 Colours of rhetoric

Although most other kinds of OSc verse contain incidental passages of rhetorical display (such as Barbour's celebrated 'Ah! fredome' passage in his Book I), sustained and concentrated use of the stylistic artifices known as colours of rhetoric (see colour n. 4) is strikingly a feature of courtly verse, particularly its set-pieces. These figures of speech were listed and prescribed in the classical and medieval treatises on poetic or rhetoric (see e.g. Atkins, 1943: 200-4; Murphy, 1974: 365-74; Ronberg, 1992: ch.5) as supplying elegant amplification. (For their copious use by Douglas, see Bawcutt, 1967: xlvi-xlviii, and 1976: 57 and 63-4.)

The simple devices favoured in oral folk-tale and in low-life verse, some of which is self-consciously in a folk-tale style (Aitken, 1978: 103ff.), only partly overlap with the gamut of high-style colours. Repetition, simultaneously verbal and of content, is not uncommon in some poems, though not as a rule sustained beyond a single re-statement. But the other devices characteristic of the 'rhetoric' of low-life verse - litotes, certain types of word-order inversion, and frequent recourse to the narrative present tense - seem not to be common in courtly verse or other solemn kinds of verse.

9.3.3 Unvernacular word-choice of dignified verse

In its diction as in other respects, courtly verse is literary and fairly slavishly derivative from earlier exemplars in Scots and, more especially, in PreStE. It is by intention unvernacular, directed towards elegant and ornamental expression free from the banal associations of daily speech. So this kind of verse, in particular, and also other sorts of non-narrative serious verse, are notable as displaying very low incidences of northern or peculiarly Scottish words, which evidently were avoided as inelegant or barbarous. The 189 lines of Dunbar's The Thrissill and the Rois, which is not untypical of its kind, contain only seven more or less exclusively Scottish or northern words, if one includes the onomatopoeic hapax legomenon swirk v., the legal compeir v. and the poetic garth n. and to-forrow adv., as well as cluvis 'paws', dully, and skaith. Even those few northernisms that do occur in this sort of poem are clustered within the brief anti-aureate passages that intentionally point up the typical courtly style by contrast.

9.3.4 Courtly diction: latinate

The unvernacular character of courtly and other non-narrative serious verse results not only from the avoidance of lexical northernisms, but also from the frequent employment of many comparatively recent word-borrowings of Latin (or Latin/French - see below) origin (latinisms) and French origin (Gallicisms), the two together being called 'latinate' diction.[191] These are presumably the “heich, pithie and learnit wordis” which James VI recommended for “ane heich and learnit purpose” in his Revlis and Cautelis. Both as types and as tokens, loanwords of this sort are common in English and Scots by the 15th century: Ellenberger counts 830 latinisms (in the narrow sense) as types, contributing 2352 tokens in the Dunbar canon (of around 40,000 tokens in toto), and the figures for Henryson are similar (Ellenberger, 1977: 22). Gallicisms would doubtless supply at least as many types and tokens again. The great majority were fairly recent adoptions and most of those most favoured by our poets date from Chaucer's time onwards, including not a few which first appear in Henryson and Dunbar (over 85 according to Ellenberger, 1977: 150).

Both referentially and morphologically these words were restricted in range. Almost all of them refer to the less basic and less general notions in their particular semantic fields: for instance, Dunbar's kinship and family terms from this source consist of four items, genetrice, matern(e, matrimony(e, successioun (Ellenberger, 1977: 49). Whereas French supplied a few conjunctions and prepositions (e.g. except, maugré, suppose), the Latin loans are exclusively in the principal word-classes. They include a few borrowings of Latin stems such as dulce, glore, laud, volt; and a few unchanged latinisms such as dirige, limbus, requiem; but the overwhelming majority are polysyllabic derived forms, e.g. in com-, con-, pro-, etc., and -abill, -all, -ance, -at, -ence, -ent, -ene, -ine, -ive, -ioun, -atioun, -ude, -our, -ment, etc.

For the most part, these conform to the French rather than the Latin morphological shape (whether or not the corresponding word is actually recorded in MF; generally it is) e.g. -ance (F -ance, L -antia), -ité (F -ité, L -itas, -itātem), -ioun (F -ioun, L -io, -iōnem). An exception is -at (= F ), as in ornat (= orné). It is usually impossible to tell, and scarcely seems to matter, whether it was the existence of a French or a Latin etymon or just the availability of the pattern that instigated the borrowing of any particular item. Perhaps, as Ellenberger argues, more often than not it was Latin.

It seems convenient, in any case, to dub such items 'latinisms' and to associate with them, as sharing much the same stylistic distributions and connotations, Gallicisms (with or without cognates in Latin), whose origins are more indubitably Old or Middle French. These include a number of nouns with the suffix -age (F -age, L -aticum - contrast -at above), such as courage, langag(e, umbrage, vassal(l)age, visag(e. Some words of this sort exist alongside cognates directly (and usually more recently) derived from Latin, for instance delit and delitabill beside delectabill, pennance beside penitence, and riall beside regall.

Doubtless most of both these kinds of words remained markedly literary in their provenance and connotations. They occur only sparsely in the more vernacular kinds of writing (such as low-life verse), but they are profuse in the instructive verse and only less so in most of the courtly verse (although Rolland's Court of Venus is perhaps the most latinate work of all), as well as in all the more literary registers of prose (see Ellenberger, 1977: 70 for some frequencies in different kinds of writing).

It is likely that many of these words remained unfamiliar to uneducated persons - the lewit, landwart or uplandis persons with whom the poets pretended, with obvious insincerity, to class themselves in the modesty passages. Ability to use them freely and to comprehend them marked one as a member of an elite. Many such words were placed apart from the unlearned vernacular by their morphology and their polysyllabicity.[192] These properties, as well as the possibilities some offered of a range of reference beyond that of their vernacular translations, helped, no doubt, to commend them to authors of dignified or pretentious verse, no less than the rhyming convenience that was presented by their limited range of suffixes. So it is unsurprising to find the courtly verse in the grand manner employing these words on principle whenever they are available.

In addition to the general body of latinate words common to the courtly poems and to other latinate kinds of writing (such as didactic verse and prose), the courtly poems also employed a special stock of highly recurrent latinate expressions, predominantly epithets, belonging to a limited range of semantic fields, on which they drew copiously and repetitively for their conventional set pieces (the locus amoenus descriptions, panegyrics, and passages in praise of the masters of poesy), e.g. angelical, auroral, celestial(l, celicall, etheriall, imperiall, nocturnall, palestrall, regal(e, terrestriall, triumphal(l, virginal - available to rhyme with unvocalised forms of words such as all (contrast rhymes dependent on l-vocalised forms, characteristic of low-life poetry); incomparabill; aureat, deificat(e, laureat, mellifluat, ornat, purpurat; clarifyit, depurit, polit(e, sugurit (s.v. succarit); eloquent, eloquence, indeficient, orient(e, redolent, resplendent, radiant, reverend, reverence; precellent, precelling , preclare; illuster; glorious, radi(o)us; cristalline, divine, matutine; regine, rosine; nutritive, restorative; amene, dulce, facund(e; rethor, rethorik; celsitud(e, mansuetud(e, pulc(h)ritud(e; diademe, paradis(e; habitakle, signakle; lucern(e, matern(e, superb(e; odour(e, vapour; decore, illumine, compile. Accompanying these is a similar body of Gallicisms (mostly di- or polysyllabic), out of courtly medieval French literature, including that of the Grands Rhétoriqueurs (see e.g. Bawcutt, 1967: xxxv-xxxvi), but mostly mediated by the Middle English courtly poets, e.g. bening, gentil(l, depaint, plesance, plesand, polist, tendir, nobil(l, rial(l; countenance, portratur(e; cheval(l)ere , genetrice, imperatrice, victrice, salvatrice; chevalry, gentillesse, gentrice, prowes, riches; grace, mercy, glore, glory, licour(e, ordour; fluris, flurist, anamal, anamaling, attemperat(e, reconfort, revest, endite.

Even more than the general body of latinate vocabulary, the Scots poets derive this stereotyped body of diction from their chosen poetic mentors, especially Chaucer and Lydgate.

Though many of the items discussed in this section were metaphors whose original reference was concrete, perhaps only the most classically minded authors and readers were fully aware of this, so they were less immediate and sensuous in their connotations than their everyday vernacular equivalents, such as gilt or giltin beside aureat, or hony-swete beside mellifluat. Equally, referring to the sun rhetorically as Phebus, Titan or Apollo, or to the winds as Eolus, and so on, was another means of exoticising the commonplace. The typical diction of low-life poetry was employed for exactly the reverse effects (see below).

It is the diction just described that modern scholars chiefly recognise as 'aureate terms' (the makars' 'termes aureat', 'poleit termis', 'facound wordis', etc.) (see Ellenberger, 1977: 82-4). Whether the makars themselves would have so restricted the reference of these expressions, or would have taken in also, say, rhetorical colours and native courtly diction, seems moot (see further Zettersten, 1979). Nor is it apparent whether they would have included the corresponding (much less extensive) body of diction consecrated to the conventional winter description, e.g. boreal(l, bustuous, penetrative (and penetrive), perturb, poleartik (s.v. Artik), pungitive, sabill, tempest(e (see further Bawcutt, 1976: 64-5).

9.3.5 Courtly diction: native

The poets also draw (especially in the descriptiones loci amoeni) on a traditional poetic diction of predominantly native origin. It is partly for this reason that the courtly poems overall score a little lower for frequency of latinisms in Ellenberger's estimates (1977: 77 and 66-9) than do some straightforward moralising poems such as 'Dunbar at Oxinfurd'. Examples include: besene, garth, gent, glete, gletering (s.v. glittering), hals(e v.2, hew n.1, lake, leme, med(e n.3, s(c)hene, strand, vale, wed(e n.1.

The same passages draw on a much lengthier list of non-latinate words of less restricted distribution, which had nevertheless by long tradition, out of earlier English poetry, become a regular part of the verdant, summery countryside scenes in verse, e.g. balme n., balmit (s.v. balme v.), balmy, bank, beuch, blome, blomit, blossom, bruke, cler(e, fair, flete, fresch(e, glaid, herb(e, hevinly, lusty, mirthful(l, soft, staneris, swet(e; and several referring specifically to brightness: beme, gleme, glans, s(c)hine, stern(e. The effect of variegated light and brightness is heightened with a profusion of names (non-latinate and latinate) of flowers, jewels and precious stones, and of colours, e.g. flour, flour delice, garland, lillie, ros(e; beriall, charbunkill, cristall, emerant, jem, gilt, gold, goldin, perle, ruby, silver, topas; blew, colour, gulis, grene, red(e, purpur(e, quhite, etc. Many of these words conform to the phonaesthetic requirements of these passages and supply suitable rhymes for them. Many figure as constituents of the conventional formulae of which the favoured descriptive passages were a tissue - the fair firthis, grene meidis, cristal knoppis, perly droppis, lemand beriall droppis, silvir schouris, mery foulis, mirthfull morowis, Phebus bemis schene, goldin skyis of the orient, blossom upon spray, dewis donk, stanniris clere as sterne, etc. (For an idea of the highly traditional character of this phraseology, see, for example, the notes to Dunbar's The Goldyn Targe in the STS edition or Bawcutt's 1998 edition.)

The winter scenes too have their more limited body of favoured words and formulae: blast, donk, daill, sleit, snaw, and penetrative air, frostis penetrive, pungitive wedder, the boustious blasts of austeir Eolus, mystie vapouris, stalwart stormis, Florais dule wede, Priapus’ gardingis bair and stormy weid.

Courtly verse favours the auxiliary do, whereas the several narrative modes favour gan, can, couth, coud (not normally followed by to, except in Lancelot), e.g. “scho tuke hir leif and furth can ga” (Henryson Fables 353), “Thay fand the town and in blythlie couth gang” ibid. 259) (see §7.8.15).

9.3.6 Vernacular diction and vulgarisms of low-life verse

The most obvious thing about the diction of the low-life verse is that it is much the most densely Scottish of any kind of writing in OSc. In part, this follows from what these poems are about. Many of them relate how certain grotesque or rustic characters have preposterously far-fetched or farcical adventures or behave in a boorish or clumsy or uninhibited way, and do this in a homely parochial setting amid everyday objects, livestock and fauna. The poems of vituperation specify directly, in a series of insulting invocations, declamations or descriptive narratives, various repulsive or ridiculous personal traits of the person addressed or described. Since these homely or undignified topics were presumably infrequent in most of the English and other literatures known to the Scottish poets, the only known terminology for them was native, local and colloquial.

More or less by definition this was the kind of vocabulary that contained the highest proportion of northernisms, including items confined to Scots, e.g. gusis cro, bowkaill stok (King Berdok); tuchet, gowk, baird, thevisnek (the low-life interlude in the Howlat). These passages often seem deliberately contrived so as to get in as many as possible of these northern terms for domestic objects: King Berdok's choice of topics seems to illustrate this, for example.

Another semantic class of words very copiously represented in these poems express disapproval or hostility or denote loud noise and violent or ungainly action. Naturally a good many are exclusively Scots.

The following examples of northernisms, mainly from Dunbar's Flyting, illustrate the character of this diction:

  • from OE: ble, derch, dreg, modwart, plat n.1, rerde, swaittis, (upon) wry; elriche, haw adj., holkit, sweir; flyte, rare, skyte, thraw, threpe, wary;
  • from ON: craig, gate, gett, carling, lisk, lug, mauch, nowt, skeil, smaik, tedder, waith; bla, blaiknit, boun, ug(sum); flyre, host, rame, rowp, rug, skar, skirl, traik; gar;
  • from Middle Dutch: cute, dok, dub, gek, loun, scaff; swanky;
  • from (? spoken) Old French: aver, barret, botine, cummer, Mahoun, grunȝe, lunȝe, menȝe; bribour and bribry, pelour, trumpour; brangill, skowder, syle;
  • from Gaelic: baird, cabrach, catherene, cryne, glen.

These co-occur, of course, with many other words of the same origins that are not confined to (nME and) Scots.

In addition to such items of known origin, a strikingly large proportion of the words favoured by low-life and flyting verse have been written off by the etymologists as of unknown or uncertain origin. Examples (again chiefly from Dunbar's Flyting) include:

  • nouns: boy, choll, clod, crele, cufe, dowsy, gane, gild, glar, gully, irle, larbar, limmer, lokman, nagus, rehatour, scarth, skill, skyre, skrumple, smaik, smy, tod, tramort, wirling; ladry, limmery; caribald, haggirbald, haschbald, luschbald; averill, gruntill; bumbard, dastard, dowbart, scutard; duddroun, ladroun, wilroun;
  • verbs: clasch, glowr, goif, gowk, lounge, roy, scale, skirl, swap, ȝou1; hirkill, hirpill, jingill, rattill, wraggill; bikker, clatter, scunner, skomer;
  • adjectives: glunschoch, harth, queir; limmerfull; bony, gend; gowkit, gukkit, glaikit; swappit; skolderit.

It is a reasonable conjecture that many of these are medieval coinages in folk-speech, and indeed some show recognisable processes of derivation, e.g. the gl-phonaestheme, and the frequentative endings -ill and -er. Several nouns and verbs seem to be echoic or onomatopoeic in origin, e.g. clasch, roy, skirl, swap, clatter, rattill. The suffixes -ard, -bald, -roun and -it (and ? -bard) form abusive descriptive terms, and -ry forms pejorative collective nouns, such as harlotry, ladry, limmery, lounry. Some items are simply specialised applications of existing words.

The flytings and lampoons include many novel abusive compounds, e.g. byt-buttoun, crawdoun (perhaps also on a verb-noun pattern), hurlebehind, lik-schilling, nipcaik, rak-sauch; girnall-ryvar, muttoun-dryvar; gallow-breid, purs-pyke, tramort, widdefow, chitterlilling, wallidraggil.

The embalmed phrases of the courtly descriptions are matched in the low-life verse by formulae and proverbial clichés, no doubt out of everyday informal parlance, such as to lauch one's hairt sair, to get one's paikis, to mak biggingis bair or to mak waistie wanis ('to impoverish oneself'), to brek someone's gall ('to break his spirit'), he had na will to mow, it was na mowis, the gallowis gaipis (for someone), quhat man settis by (one's adversary)?, quhat or quhare devill?, (one's adversary or butt is some notoriously disreputable individual's) air, (to sink in something) up to the ene.

Many of the words and phrases distinctive of the low-life poetry are special in their lexicographical histories, appearing fleetingly or intermittently on record. Of those whose origins are known, a high proportion (e.g., from ON, carling, lug and smaik) are (all but) unrecorded between the source language and their emergence in the low-life Scots poetry. The same is true even of several of the words of OE origin, such as elriche, haw, plat, swaittis. A still higher proportion make their first or almost their first appearance in the Scots poetry. Some of these have a later history in English, though often, as with lounge, queir and up to the ene, only after an interval of a century or more. Some do have a continuous history in Scots. Others are ephemera that fail to outlast the 16th century, e.g. gane, gend, larbar, smy.

Many of these words are rare even in OSc. Some are hapaxes or occur only twice or three times in the low-life poetry only. Others are found also in a quite restricted set of other contexts: violent or condemnatory passages in other verse and prose, and specimens of alleged 'flyting and bairdrie' cited in court records (for examples, see Cusack, 1998: 15ff.), and in certain works of narrative prose in late MSc, beginning with Knox's History.

The discontinuous lexicographical history and general rarity of some of these words can be simply explained from the fact that the detailed attention to domestic operations, intimate personal characteristics and physical traits given in this class of poem does not happen in any other kind of writing: an example is the process of butter-making described in The Wyf of Auchtirmwchty but nowhere else in MSc, so that the terms bledoch and ȝyrn occur only there. But the rarity of other words of this diction has a less trivial explanation. They belonged to an essentially colloquial or slang register and so were appropriate only in writings that imitated this. Their emergence into the limelight of literature had to await the appearance of this copious body of writings.

It will not surprise the reader that the latinate diction and native amene vocabulary are almost totally absent from low-life poetry. But, as befits their educated contestants, a small proportion of latinisms and gallicisms is present in the flytings and lampoons.

The alliterative native diction also occasionally supplies a few items to low-life and flyting verse, e.g. synonyms for 'man' (such as berne, freke, sege), and a few tags such as on raw (all of these, for example, in Dunbar's Flyting). In the tournaments, the 'fair and fracas' poems, and the burlesques, this is sometimes presumably mock-heroic or by way of parody.

9.3.7 Reduced forms

The low-life poems do not display any of the anglicised forms described above, and their copyists, sensitive to stylistic proprieties, normally impose none on them (with the sole exceptions of the all-pervasive spellings no and so). Low-life poetry does, nevertheless, have its own formal stylistic markers. These are the written reflections of reductions of unstressed words in rapid speech, and of recent innovations in pronunciation involving the vocalisation or loss of consonants. All were probably still only optional in speech, existing alongside alternative full-form options, as indeed most continue to do in ModSc today. Specifically, these are:[193]

  • forms resulting from l-vocalisation (see §6.23);
  • forms resulting from the loss of intervocalic and word-final /v/ (see §6.31.5);
  • forms due to loss of /θ/ (see §6.31.5);
  • others similarly losing final /f/ or /v/: -sell (beside -self) as in himsell, thairsell(is), ȝoursell; twell (beside twelf);
  • en and sen beside end and send;
  • beid, dude, ford, kend, etc. for be it, do it, for it, ken it, etc. (see 'd);
  • apparently confined to dialogue are reduced pronoun-verb operator phrases: Is, weis, ȝeis for I sall, we sall, ȝe sall (see S2) and Ile, ȝele for I will, ȝe will.

In his Revlis and Cautelis, James VI specifies “cuttit short” forms such as “Iis neir cair” (for “I sall neuer cair”) as specially appropriate for “flyting and inuectiues” and adds that in “loue or tragedies” “ȝour wordis man be drawin lang quhilkis in flyting man be short”. In rhyme these reduced forms are indeed frequent only in comic and satiric verse (six of The Wyf of Auchtirmwchty's 60 rhymes are reduced forms), and in certain narrative poems, such as Hary's Wallace, the Asloan MS Sevyne Sages (see van Buuren-Veenenbos, 1982: 98-101 and 123), Douglas's Aeneid, Stewart's Chronicle and Rolland's Seaven Seages; but in serious lyric or didactic verse and in courtly narrative they are rare or absent.

The few rhymes requiring reduced forms of this sort in Douglas's The Palice of Honour, all occur in what we might accept as appropriate contexts, including irritation and contempt (e.g. stupifak, line 1460). The same explanation does not seem to apply to the substantially larger number of such rhymes in John Rolland's The Court of Venus. Perhaps we should regard this as an extension by Rolland of a licence that he imagined he had observed in Douglas's poem.

Sometimes the copyist presents us with a full-form spelling where the rhyme demands the reduced form (there are examples in Kynd Kittok line 13 (Ch. & M.) and The Wyf of Auchtirmwchty lines 9, 10, 33, 50 (Bann.)).

As spellings these forms seem to occur only in the same poems. They are also a prominent feature of the colloquial Scottish speech of the Jockies and Jamies of some English dramas of the Elizabethan and Jacobean period, e.g. Ays or I'se (= I shall), thouse (= thou shalt), I'll (= I will), deel (= devil) and our sell (= ourself). Their incidence in MSc prose requires further study, but they do appear to be much commoner in less conventionally spelled texts (such as some ill-spelled private letters of the 16th century).

When the OSc poems were written these forms were comparatively recent innovations in speech. They remained unacceptable - presumably as colloquial modernisms - in general written usage and emerged into regular use only with the ModSc dialect verse of Allan Ramsay and his followers (including Robert Burns) in the 18th century.

9.3.8 Stylistic opposites

It is possible to compile a list of synonymous alternatives in MSc, one (or more) of each set being favoured by serious prose and dignified verse, and another by more vernacular kinds of writing. In pairs like the following, for example, the first appears to have had more solemn overtones: know and ken, ere and lug, fox and tod, hound and dog (or, more abusively, tyke). Old French supplies the more dignified alternative in the sets pas and gang, promise and hecht, requeist and ask or speir. A particularly revealing set is that of the synonyms for 'face': the dignified (and latinate) visage or countenance, the common core item (of Old French origin) face, and the vernacular (and more or less abusive) front, gane, gruntill, grunȝe, snout. The amene redolence, odour sweit or dolce odour of courtly poetry likewise has its low-life verse opposite in such expressions as foul stink.

9.4 Syntax

9.4.1 Verse (Aitken, 1983)

Aitken (1983: 31) judges that no variety of Older Scots verse compares for average syntactic complexity with the most syntactically elaborate kinds of prose, but the less vernacular styles of verse are nevertheless characterised by greater syntactic complexity (see below).

More vernacular styles favour simple sentences in parataxis (i.e. in sequence, without connectives), or co-ordinated by the emptiest connectives, and with frequent asyndeton (omission of conjunctions of co-ordination and subordination and of relative pronouns, and some other types of ellipsis) and occasional parenthesis. Parataxis and asyndeton (including ellipsis from the sentence of initial words of low information content) are common also in the alliterative narrative verse such as Holland's Howlat (e.g. at lines 497ff.). The extreme of this tendency is sustained with high consistency in the narrative of Hary's Wallace.

9.4.2 Oral narrative (Aitken, 1978)

Aitken (1978) describes the style of three texts derived from or imitative of oral narrative, characterised by low-style diction and cuttit short forms of words, as well as by colourful idioms and distinctive syntax. These are ‘John Campbell's Complaint’, from the Lanark Town Council Records, c1613; a passage from The Pockmanty Preaching (J. Row, Sermon); and the Wyf of Auchtirmwchty (Bann. MS, STS edn., pp.320-24).

The syntax of these texts is relatively simple, and the subordinate clauses that are used are often parenthetic:

(that is, with the grammatical subject expressed within the clause itself so that there is no overt structural relationship beyond the clause, as he beand lodgit in Bessie Wilkeines hous) … … one property of this sort of clause is to allow the author to leave unexpressed and often unclear and perhaps undecided the exact relationship of the action of the subordinate clause to that of the main clause. (p.100)

However, there are various focussing devices, some of them grammatical constructions that are common in modern speech, but otherwise rare in OSc:

  • fronting of an adverb of motion (see §7.11),
  • the historic present tense (see §7.8.2),
  • the Who was there but so-and-so construction,
  • ellipsis of verbs of motion (see §7.8.17),
  • fronting of a noun phrase with the help of a recapitulatory pronoun, e.g.:
    The twa gaislingis the gled had left,/ That straik dang baith thair harnis owt (Wyf of Auchtirmwchty ll.79-80)
  • the redundant use of anticipatory or recapitulatory pronouns:
    The first that he gat in his armis,/ It was all dirt vp to the Ene (ibid. ll.89-90).

9.5 Prose

Older Scots prose was slow to develop and sparse in its achievements. There was ... little in the way of religious prose; no Malory emerged to create a Scots prose romance ... The historians almost all wrote in Latin ... or in vernacular verse. Significantly, when in the mid fifteenth century, an unknown Scot came to translate the French prose Lancelot into Scots, he wrote in pentameter couplets. Only in the second half of the sixteenth century did vernacular historians and religious controversialists emerge in any numbers ... Not until the middle of the seventeenth century, and then in English, do we find, in Sir Thomas Urquhart's Jewel, a Scottish writer attempting imaginative, fictional prose ... (Lyall, 1988: 163)

In England, fifteenth-century translators of French romances … had begun to use prose; it is in these translations … that English prose fiction begins. Why their Scottish counterparts did not, so far as we can tell, do likewise is a mystery … (Riddy, 1988: 46)

The relative lack of literary prose in Older Scots must be seen in the context of a learned scholarly elite, educated in Latin, who addressed a wide international audience in Latin; while for light reading, and even in most cases for the narrative entertainment provided by history, the preferred medium was verse.

9.5.1 Genre

Meurman-Solin (1993a) offers a classification of prose text-types, and also a selection of extracts. (For some further colloquial texts, see Cusack, 1998.) Meurman-Solin describes the genres in some linguistic detail (1993a: 86ff.), but the classification is based "solely on extralinguistic criteria" (p.148), and her findings mainly throw light on processes of sociolinguistic change and genre-governed variability, rather than literary stylistics. There are large variations between and within texts in most categories, as she acknowledges. Her study of linguistic features in terms of the dimension 'Involved' (i.e. addressee-oriented) versus 'Informational' (1993a: 276ff.) allows her to characterise the genres in linguistic terms, and to explain some of the internal variation within genres. Her (1993b) treatment adds the dimension of 'evidentiality' (i.e. comments on truth-values), and measures this together with the factors 'addressee-oriented' and 'author-centred'. These studies confirm, for instance, that diaries are very variable in the extent to which they are 'author-centred' (e.g. using the first person singular pronoun at a relatively high frequency), this being most characteristic of the diaries that focus on the inner religious life of the writer.

9.5.2 Style

It is difficult to generalise about stylistic trends amongst the isolated outcroppings of literary prose in ESc and early MSc, particularly as much of the work is translation, heavily and idiosyncratically influenced by originals in other languages. The introductions to STS editions of individual texts generally discuss the language and style of the work in question. There is also a small number of other linguistic or stylistic studies: Sheppard (1936) and Farrow (1997) on Bellenden's Boece; Jack (1981) on Knox; Ledesma (1996) on The Complaynt of Scotlande, Jumpertz-Schwab (1998) on Bellenden's Boece and Livy, and Dalrymple's Historie, with some attention also to other prose writers. Corbett (1999) deals with a wide range of prose translations. Lyall (1988) and Reid (1988) together provide an overview of OSc prose.

From the starting point of stylistic effect (as opposed to genre), the dimensions of language that have attracted most notice are:

  • latinate versus vernacular diction,
  • syntactic complexity (which goes together with latinate syntax) versus 'plain' writing,
  • and, in the case of translations, the influence of the original language.

No-one has yet attempted the kind of holistic characterisation of the stylistics of OSc prose offered by Aitken (1983) for OSc verse, but his observations on the grammar and diction of literary prose (1971) make a useful starting point:

It is … possible to detail certain common tendencies and usages … which characterise, at least in their most dignified passages, literary prose, the principal records and most other official writings, and some of them also serious verse: a general tendency to hypotactic rather than paratactic structure and towards the use of certain constructions, such as the … 'accusative plus infinitive' construction, which were perhaps not normal to unstudied vernacular speech; a fondness for passive and impersonal constructions; a free use of loanwords of Latin or literary French origin in addition to, or instead of, equivalent vernacular expressions; in the later writings, the sporadic use of English-derived forms as alternative to their native cognates; the employment of quh forms of the relative where vernacular usage more probably had that or the 'zero-form'; the (more or less sporadic) habit of inflecting certain adjectives in plural concord; and the avoidance of certain constructions and of certain words which apparently had specially strong colloquial overtones [see §§9.3.6, 9.3.8]. (pp.177-8)

On the formal Scots of legal and administrative prose, see §8.3.3.

9.5.2.1 Latinate tendencies

The use of latinate diction is individually variable, and partly dependent on subject matter.[194] Jumpertz-Schwab, following Weinreich (1968), lists the main reasons for lexical borrowing: the need to designate new concepts, the feeling that some of the semantic fields of the borrowing language are insufficiently differentiated, and the association of positive social values with the source language. The last mentioned, she concludes, accounts for the larding of Middle Scots texts with French and Latin borrowings, many of which have not survived to the present time (1998: 15).

Syntactic complexity is, in varying degrees, typical of most OSc prose, both the carefully convoluted prose of texts with legal status, and the grave ponderousness of texts aspiring to authority. Syntactic complexity can, however, be an elusive concept. The term is often used loosely by critics to convey a general impression rather than to pinpoint specific features of syntax. Here, following Leith (1983: 110-13), we can list three ways in which the syntax of a piece of writing can be said to be complex:

(a) heavy use of subordination as a means of linking clauses, especially when there are subordinate clauses within subordinate clauses. A case of the latter is the relative clause with an embedded temporal clause, described and exemplified by Jumpertz-Schwab (1998: 192ff.):

ane Ile … Quhair quhen they cam … thay fand it mekle bettir than the rumour was;

(b) the placement of subordinate clauses or other lengthy optional structures, in such a way as to interrupt the flow of the sense.[195] In general, adverbials are easier to process if they come at the end of a clause, after all the main structures (subject, verb, object, complement). In a tree diagram, the movement is right-branching, with levels of subordination moving rightwards and downwards (see Quirk et al., 1972: §11.81). However, scene-setting adverbials of time and place are naturally positioned at the beginning of the sentence, as are conditional and concessive clauses (e.g. those introduced by gif or althoch);

(c) complexity of tense and aspect in the verb phrase (see §7.8.18).

All of these tendencies can be traced to the influence of Latin and ultimately Greek. Blatt recalls

Aristotle's distinction between the 'strung-together style', "which has no end in itself, unless the subject matter comes to an end" and the 'periodic style'. (1957: 36)

the latter having an architectural quality:

It makes use of a fine system of subordination which allows the possibility of grouping a very large number of facts, indicating their mutual relations, in one syntactical unity, that is to say a sentence which may be long but nevertheless is very clear … … There is no need of grouping and classifying facts linguistically, if you have no complicated thought to express. That is why the art of building a period belongs more to written language and to carefully prepared speech … (ibid.)

As we saw in §7.8.18 and §7.13, many specific constructions were borrowed from Latin (directly or via French) or were developed under the influence of Latin, in order to emulate the eloquence of the classical languages.

9.5.2.2 Anti-latinate tendencies

Two tendencies pull in directions other than the latinate. One is the functional, which favours straightforward right-branching syntax (see above), and comes into play when no literary aspirations or bureaucratic training intervene, as in private letters; and also comes into play, at its most extreme, in the relatively rare texts derived from speech, such as direct speech quoted in trials, and extempore preaching (see §8.3.5.3).

The other anti-latinate tendency is the Biblical, found for instance in sermons and some religious instruction. This has been little considered in the description of OSc prose stylistics, and deserves further research.[196] Critics, echoing contemporary terminology, generally perceive non-latinate syntax (such as that of Gau, translating from the Danish) as plain (q.v. adj.1 7), but in practice it is often characterised by parallelism, pairing of near-synonyms, and (often as a consequence) rhythmical phrasing. We find this high vernacular style also in such texts as a formal cursing ("Excommunication, in the vernacular, of Border reivers, 1525", in St. A. Formulare) and prose prayers (see Bennett, 1955; Aitken, 1957), rare survivals of whole genres wiped out by the Reformation. Ironically, given the importance of Biblical translation into vernacular languages, the high vernacular style is not especially a phenomenon of the Protestant Reformation, and indeed the writers of the Catholic Tractates were critical of the 'curious and affected' style of some of the Reformers (see plain adj.1 7).

These Biblical tendencies - and no doubt others - resonate with the elevated styles of folk literature (the ballads or muckle sangs, and presumably also, as in ModSc, rehearsed passages of storytelling). The emphasis on anglicisation as a kind of national betrayal, and the association of StE in modern sociolinguistic research with status and prestige, may lead us to misrepresent the significance of StE to the average Scots speaker in the period before universal education. The position of Biblical English as a bridge between elevated styles of Scots and a certain kind of elevated English deserves further attention.

[175] This sub-section (§9.1) is a revised version of part of Macafee (1988).

[176] Classified in the 19th century by Sievers as five types: see any standard work on OE, such as Campbell (1959).

[177] Dunbar's The Twa Mariit Wemen and the Wedo is the only important poem in which the unrhymed alliterative line is used in Scots.

[178] Phonetically, /h/ is the voiceless equivalent of whatever vowel follows it.

[179] Already in late OE an alternative 'loose alliterative verse' form had emerged (Turville-Petre, 1977: 8). It has been suggested (Crépin, 1978) that the more analytic grammar of the later language made for longer clauses than in Old English (though we must remember that many endings were lost).

[180] Riddy (1988) makes the point that serious poetry in Scots with "marked alliteration" was still being written in the first decades of the 17th century, but she does not distinguish between the structural alliteration of pure stress metre and the optional alliteration of stress-syllable metre.

[181] But already found in ME in The Awntyrs of Arthure.

[182] Kirkpatrick (1978) suggests that the thirteen-line stanza was originally a dance form, and the bob provided a beat on which a circle of dancers could change direction.

[183] In Meurman-Solin's (1997b) data, the spelling –t without the vowel occurs first in verbs ending with –is (1450-1500).

[184] Dunbar's "Done is a battell on the dragon blak" is an interesting case of a poem in which almost every possible metrical licence is exploited.

[185] On stanza forms, see Corbett (1997: ch.12 and Appendix) - CM.

[186] Priscilla Bawcutt (personal communication) disagrees with the placement here of Bannatyne's first section "which contains religious poems, whose style is very varied” – CM.

[187] Cf. Nichols (1931). Bawcutt suggests that, "… it might better be termed 'International Late Gothic'. It was widely fashionable in the late 15th century and practised not only by English poets but by many French and Burgundian ones" (1992: 358) - CM.

[188] Or in some cases, such as the inflected plural forms of verbs, were - CM.

[189] On Dunbar, Bawcutt writes, "It is important … to see such anglicisms in perspective. They occur only in a small cluster of Dunbar's poems … If one examines Dunbar's usage as a whole, what is most striking is the comparative sparseness of anglicisms; this contrasts with the remarkably profuse and eclectic use of such forms by Douglas" (1992: 359) - CM.

[190] The following verse from Lydgate's Right as a Ram's Horn as found in the Bannatyne MS is typical of the form in which the English poets were read in Scotland. The features that were adopted were necessarily of the kind that would show through in the versions read.

All rychtouß thing the quhilk dois now proceid
Is crownit lyk vnto an empereß
Law hes defyit guerdoun and his meid
Settis hir trewt on hicht as goddeß
Gud faith hes flyttin wt fraud & dowbilneß
and prvdence seis all thingis þat cumis beforne
ffollowing the trace of perfyte stabilneß
als evin be lyne ryt as a rammis horne.
(Bann. fol.79v)

For further discussion, see Boffey and Edwards (2000) - CM.

[191] Meurman-Solin regards as particularly marked those loans that represent the first appearance in the language of that stem (1993: 231ff.) - CM.

[192] Meurman-Solin finds that "marked loans are seldom used in simple phrases which function as immediate sentence constituents" (1993: 233) - CM.

[193] Most of the following examples and some of the subsequent discussion are from Aitken (1971) - CM.

[194] See for instance Jumpertz-Schwab's (1998) comparison of Bellenden's translations of Livy and of Boece.

[195] Cf. Aitken's use of the traditional term 'hypotaxis', referring to clauses “in which the noun-phrase and verb-phrase elements of sentences are modified by words, phrases and clauses” (1983: 31).

[196] But cf. Jack (1971: 24) who considers it to be mediated by sermons.

Macafee, Caroline and †Aitken, A. J. (2002) ‘A history of Scots to 1700’ in A Dictionary of the Older Scottish Tongue vol. XII, xxix-clvii. Online http://dsl.ac.uk/about-scots/history-of-scots/style/