ME syntax: more than Pyles and Algeo want to tell you
Principal source: for these notes
-Fischer’s chapter on “Syntax” in Volume II: 1066-1476 of the Cambridge History of the English Language (Cambridge UP, 1992).
-another source that I haven’t confirmed yet, possibly Lehmann on Indo-European
Examples and supplementary points from:
-Closs Traugott’s chapter on “Syntax” in Volume I of CHEL (Cambridge UP, 1992)
-Mustanoja’s Middle English Syntax (1960)
-Oxford English Dictionary
Big development from OE to ME:
- from a more synthetic language (relying on inflections) to an analytic one (relying on relatively fixed word order, schematized as SVO)
- from a language “which was largely verb final ... into one that is clearly verb non-final”
Still some exceptions to SVO word order in PDE:
- emphasis: “Fish, they will eat.” (O S AUX V)
- specific constructions, e.g. VS inversion
- after negatives or implied negatives: “Never have I taken such a useless course.”
- in questions: “Is the end in sight?” (VS)
Questions to consider:
Briefly, review the dilemma about OE
- verb-final?
- SVO with relics of verb-final?
A bit less briefly, can we determine a relationship between
- fixing of word order
- loss of inflections
More extensively, what are the implications or concomitants of these changes? Some are
- proliferation of prepositions
- ex: OE verbs that took objects in the genitive case by ME have prepositions rather than case endings (but sometimes no prepositions)
- increase in verb phrases of various kinds
- ex: by late OE, indicative and subjunctive become indistinguishable in many slots, and we find more examples of the ‘periphrastic’ subjunctive with what were to become the modal auxiliaries (should, may, etc.)
- changes in the order of other elements
- ex: from preposed to postposed possessive: OE his feonda slege -> the slaughter of his enemies
‘Linguistic universals’: Entrenchment of VO order should entail other changes
Greenberg’s ‘linguistic universals’
- e.g. “if a language has front rounded vowels, then it also has front unrounded vowels”
- but the best-known ones relate to word order
- “Word order typology revolves around various permutations of the magical letters S,V, and O” (Watkins)
- and on the order of elements within the noun phrase and the verb phrase
- e.g. order of adjectives relative to nouns
- e.g. order of possessives relative to nouns
- e.g. order of nouns relative to adpositions (cover term for prepositions)
OV languages tend to have
-prepositions that follow their objects (“herein”)
-adjectives that precede their noun (“the fat cat”)
so, English has maintained the ‘OV’ pattern for descriptive adjectives
VO languages tend to have
-prepositions that precede their objects (“in here”)
-possessives that follow their noun (“the ”)
Why?
- innate? is there a switch in the brain for “head first” vs “head last”?
- perceptual: inconsistent languages would have centre-embedded train-wreck structures that take up too much short term memory. We can handle
- “The cheese the rat the cat chased ate was rotten.”
Why not?
- Although language universals are ‘statistically significant’, consistent languages don’t seem to form a strong majority
- “It is because the lampposts are there that the drunks are seen to lurch in one direction or another” (Matthews)
- “A language in transition [from OV to VO] is like a drunk moving unsteadily between two lampposts. Latin, for example, had somehow lost its grip on the OV lamp-post, and, in acquiring prepositions, had taken one step towards the other side of the road. Later, over the centuries, it took others ... in this way, the typological laws explain particular historical changes, or series of changes. It is because the lampposts are there that the drunks are seen to lurch ... in one direction or another.” ... “If inconsistencies remain, then “they are due to social forces, pragmatic actors, and above all to borrowing. When all these variables are excluded there is always the ultimate escape-hatch that language X has not undergone the change yet.”
Word order before OE
(from Lass, Old English)
Word order in PIE and PGmc very contentious! Why?
- because there is no written evidence has to be ‘reconstructed’
- and reconstructing syntax (order of words) not the same as reconstructing phonology and morphology (within single words)
- assumed to be “basically SOV” “though there is of course considerable debate”
- interesting that many descendants of PIE have SVO...
Word order in NWGmc basically SOV
- “small and often obscure” runic corpus from 3rd-7th century AD
- transitive clauses: OV:VO about 4:1
- about 70% OV: “
- I Hlewagastiz Holtijaz horn made”
- less than 20% SVO:
- “I-and-my-fellows-Wiwio wrote this rune”
- [prepositional (most SOV languages are postpositional)]
- [postmodifying (“dog big”)]
Word order in OE
(from Lass, Old English; Fischer in CHEL2)
“Linguists still argue on the point whether OE was truly a SOV-type language that changed into an SVO language in the course of the ME period. The difficulty in deciding this matter lies in what kind of theoretical attitude the linguist wishes to adopt.”
- ‘mentalist’, ‘generative transformational linguist’ “will wish to look behind the surface word orders in order to find a basic or underlying word order that will explain the presence or absence of certain surface constructions in the language. The base order this linguist hypothesizes need not necessarily be the most frequently attested surface order in a language. In most cases this type of linguist has come to the conclusion that OE was verb final.”
- ‘behaviorist’, “the more taxonomically inclined linguist, however, who bases himself primarily on the percentages of SVO, SOV and other orders in both main and subclauses, would probably come to the conclusion that OE cannot be said to represent any one type because of the amount of variation present. If pressed into a choice, he might well opt for SVO.” I.e. so much variation that you can’t come off the fence.
- “but there is general agreement as to what type of language ME finally became, i.e. an almost pure SVO language.” (Fischer, CHEL2, 371).
What about early OE? Lass looks at Caedmon’s hymn (this is the Northumbrian version)
- has both OV and VO transitive clauses
- or astelidae, middungeard ... tiadae
- basically prepositional: heben til hrofe
- basically premodifying: hefaenricaes uard, haleg scepen
- but uerc uuldurfadur, frea almectig
What about early OE prose? Lass looks at one early text
- OV order declining relative to the NWGmc corpus: OV:VO about 1.5:1
What about OE prose generally?
- verb-final order gets increasingly restricted to
- subordinate clauses
- pronoun or other ‘light’ objects
- short
- carry little new information
Typical OE sentence:
And đa he hig forlćten hćfde, he eode on ţone munt, and hyne ţćr ana gebćd
And when he them released had, he went up the mountain, and [him] there alone prayed.
Word order changes in ME
Verb-final order hung on in ME in some contexts
-in different ways in the North and the South (Kroch and Taylor)
-contact with Scandinavian in the Danelaw
-don’t homogenize ME!
-with pronoun objects
-in the Ormulum
-51% of pronoun objects OV (in all clauses)
-18% of nouns
And VS inversion retained for quite a long time
-after wh- elements
“Why make ye youreself for to be lyk a fool?”
-after adverbials
-perhaps not a problem “because in all cases the subject was still next to the verb”
And (optionally) after negatives: Never have I taken such an exciting course ....
-in OE VS after ne
buton
hwa beo ge-edcenned of wćtere, and of Haligum Gaste,
unless one [who] be reborn by water, and by the holy ghost
ne mćg he in-faran
on Godes rice
not can he go into God’s kingdom
-in ME, negation shifted from ne + verb (+ naht) to verb + not
-the naht shifted from a stylistic option to being grammatically required
-and then ne could be dropped since it was supported by naht
no but a man schal be born a3en, he may not se the kyngdom of God (Wycliffite bible)
-but perhaps since many negatives were adverbials (never, scarcely) and since “the earlier system” (OE) had permitted VS after negatives, we have “a renewed grammaticalisation of the inversion rule after negatives and implied negatives” (377)
scarsly shaltou fynden any persone that may kepe conseil secrely (Chaucer)
But SVO order establishes itself quite rapidly at least in the ‘continuations’ of the Peterborough Chronicle (last entry 1154)
-Mitchell found 72% and 88% SVO order in subclauses (cf 41% in his late OE prose sample (Aelfric))
Disappearance of inflections or Fixing of word order?
Arguments in favour of ‘loss of inflections’ (nb there are other arguments for and against)
Work on OE
Dancev looked at OE phrases with and without prepositions
-Phrases without prepositions have NP elements with lots of formal distinctiveness
-Phrases with prepositions have NP elements that don’t tell you very much
-the more meaningless the inflection (i.e. –e) the more likely you are to find a preposition
-argues that the loss of inflections would have to come first
Comparative work:
Gerritsen looked at Middle Dutch and argued that loss of inflections triggered SOV -> SVO
-“statistical evidence” for “a strong and significant interrellation between development from SOV to SVO and loss of inflections”
Fischer’s summary of the implications of studies like that (p. 374):
“With the loss of inflections, the greatest need was for subjects and objects to become distinguishable since both were normally represented by NPs. Because only pronominal subjects and objects could still be distinguished in case, SOV order first disappeared in the case of nominal NPs.”
SVO order has impact on
‘Impersonal’
constructions in OE
o
difficulty
defining:
o
hit as subject
§ Hit licode Herode
o
no expressed
subject
§ Hu đincđ eow?
§ Swa mé đyncđ.
§ eow gebyrađ
In
ME an expressed subject before the verb becomes increasingly obligatory:
o
‘dummy’ (h)it,
there
§ It was pleasing to the king,
It pleased the king
§ It behoueth 3ou
§ It seems to me
o
but there are
still examples of subjectless constructions:
And happed so, they coomen in a toun
o
animate
experiencer (formerly in oblique cases) reanalysed as subject?
§ The king liked ...
§ I think
S V
o
argument over
whether the verbs changed in meaning
o
from causative,
e.g. OE lician ‘to please’
o
to receptive,
e.g. ME like ‘to like’
o
or had a more
general meaning,
o
e.g. lician
and like ‘the existence of pleasure’
And
SV order might also be correlated with the disappearance of some of the OE
‘correlative’ conjunctions
Correlatives
in OE
-many
had the same form whether they’re adverbs or conjunctions,
e.g. đaer ‘where’ SV, đaer ‘there’ VS
e.g. đa ‘when’ SV, đa ‘then’ VS
Around
in eME with SV / VS word order:
And đat ođer dei đa he lai an slep in scip,
‘when he lay asleep in the ship’
đa đestrede đe daei ouer al landes
then darkened the day over all the land
ţanne he komen ţere ţanne was Grim ded
when he came there then was Grim dead
But
as word order gets fixed as SV, we find “a more transparent system, in which
conjunctions are distinct from adverbs”:
-e.g. in Chaucer, ţo is an adverb but not a
conjunction
-when (that) starts to be used (from OE
interrogative hwanne ‘when’)
Same
text:
Wan ţe godemen ţat sawe / he
stirten up sone anon
When the worthy men saw that ... they
leapt up all at once
Word order changes relating to noun + modifier
According
to ‘language universals’ languages with VO order tend to have head-modifier
order
-noun-descriptive adjective: “dog big”
-but “English has maintained the OV pattern for
descriptive adjectives”
-possessive construction: “the dog of
his neighbour”
How
and why did the of-phrase appear in English?
o
“may have been
helped along by the parallel French construction with de”
o
some evidence
of its higher frequency “in works written under strong French influence”
o
but “a native
development, parallel with
o
Romance
developments from Latin de
o
Dutch van
and German von”
§ arguably a concomitant of
SVO word order
Of
course the inflected genitive survives, but mostly
o
in adnominal
use, with possessive/subjective function: “John’s book”
o
reinforced by
S(V)O order: “John has a book”
Where
and why was the inflected genitive otherwise lost?
o
‘objective
genitive’ more likely to go:
o
e.g. OE. his
feonda (gen.) slege ‘the slaughter of his enemies’
o
e.g. OE Iudea
ealdor ‘the leader of the Jewish people’
§ an essentially OV
structure (enemies-kill, people-lead)
o
in ME, with SVO
typology, this structure becomes ambiguous/opaque
o
exceptions:
phrases like the king’s assassination or his death
§ it’s clear who’s being
killed!
o
genitives
dependent upon adjectives,
o
e.g. OE Beođ
hyra geóca gemyndige ‘they are mindful of their safety’
§ genitive-adjective
relationship often OV
o
in ME, VO
relationship expressed with adjective + of-phrase
ME myndfule I was of dayes old
Do
we find any postmodification in ME?
-as
in OE, where two adjectives are involved: a good man and a fair
-in
adjectives borrowed from French: goodes temporels, service divine
-in
poetry on metrical grounds
-with heavy adjective phrases: wise advocatz lerned in the lawe
Prepositions in ME: more, and more often
Representative
contexts:
Verbs
which in OE took objects in the genitive,
§ OE examples, e.g. wundrian
‘wonder’, wyscan ‘wish’, e.g.
[Hie] his
tocymes wyscton.
o
though the
genitive inflection –es was a relatively healthy one, nevertheless
in ME, we find these verbs taking
o
prepositions
§ wonder took of first, but
often then lots of variation (e.g. OED has wonder at, sometimes over,
formerly of, on, upon)
§ wish could take after,
now for
c1200 Trin.
Coll. Hom. 3 Men..wisten 
erne
after ure lauerd ihesu cristes tocume.
o
sometimes only
an oblique case
§ c1470 HENRY Wallace IX. 1413 Sum
wald haiff had Boyd at the suerdis lenth; Sum wyst [ed. 1570
wissit] Lundy.
§ Many a man
wonderinge the bewtye of a straunge woman, haue bene cast out
Unlike
genitive -s, dative inflections did ultimately disappear.
What
happened to monotransitive verbs with dative objects?
§ OE examples, e.g. helpan
‘to be of assistance to’
u
mone
um
helpst.
§ in ME, the dative usually
becomes indistinguishable from the accusative
1382 WYCLIF Rev. xii. 16 The erthe helpide the womman.
What
about ditransitive verbs (e.g. offrian: acc. (d.o.) and dat (i.o.)
OE:
You offer God (gode, dat.) a sacrifice (lác, acc.)
ME:
offer God a sacrifice / offer a sacrifice to God (i.e. no dative
inflection; preposition obligatory when DO IO)
Prepositions – a digression
A
few new prepositions were borrowed, e.g.
till adopted from ON (& OE ođ
‘until’ becomes obsolete)
accordant
to,
during adapted from Lat/OF durant(e)
Some
of the new prepositions were participial: e.g. accordant / according to,
considering, during, excepting,
c1386 CHAUCER
Frankl. T. Prol. 3 And gentilly I preise wel thy wit, Quod the
ffrankeleyn, considerynge thy yowthe, So feelyngly thou spekest, sire, I
allowthe.
Did
they develop from absolute uses of participle (in concord with the subject),
into prepositions? The OED doesn’t actually find many “clear examples of
the former”, e.g. *‘considering
his youth, we were surprised at his attainments’. They assume instead
that the present participle considering was substituted for the past
participle considered “(formerly
placed before its n.), is used in an absolute clause, = ‘being taken into
account’.”
?1426 Lett. Marg. of Anjou (Camden 1863) 33 Considered
this that here is reherced.
Many
prepositions like during, notwithstanding were adapted from
Fr/Latin constructions (durant, non obstant):
The pres. pple.
of DURE v.
= enduring, lasting, continuing, was used in Fr. and Eng. in a construction
derived from the Latin ‘ablative absolute’; thus L. vita durante, OF. vie
durant, Eng. life during, while life endured or endures.
1480 CAXTON Chron. Eng. lxxxviii. 72 She neuer was
seyn among folke hir lyf durynge.
The participle also
often stood before the n., e.g. L. durante bello, F. durant la guerre,
Eng. during the war; in which construction during came in the
modern langs. to be treated as
c1385 CHAUCER L.G.W. Prol. 283 (MS. Gg. 4. 27)
Stedefaste wedewys durynge alle here lyuys.
But
many prepositions of native origin were originally adverbs
In
PDE, the same word can function as both, e.g.
They
went up / They went up here / They went up the hill
She
sat near / She sat near me.
For
example, before and behind were used adverbially in OE before
they became prepositions:
“Behind is used both absolutely (as
adv.), and with an object (as prep.), the latter originating in an OE dative of
reference, behindan him ‘in the rear as to him’.”
Here’s
the adverb be hindan: c900 O.E. Chron a
Deniscan sćton 
ćr
be hindan.
Here’s
what the OED classifies as the preposition behind: “In a place
left by (one who has gone on). Usually with leave, remain, stay,
expressed or understood.”
c1200 ORMIN
8913 He wass a
bihinndenn hemm bilefedd att te temmple.
Before: Primarily an adverb; its
relation to a n. was expressed by putting the latter in the dative, ‘in front as
to a thing,’ whence it passed into a preposition (cf. B 2, quot. 971).”
[971 Blickl.
Hom. 15 [He] ehyrde
myccle menio
him be~foran feran.]
c1000 ĆLFRIC
Ex. xiii. 21 And Drihten fór beforan him and swutelode him one
we.
And
many prepositions (before that) can also be conjunctions (via that-clauses?). e.g. ‘think before (that) you speak’
c1200 ORMIN 964 Biforenn att
te Laferrd Crist Wass borenn her to manne.
c1325 E.E.
Allit. P. A. 529 On oure byfore e
sonne go doun.
c1386 CHAUCER Frankl. T. 233 By cause that he
was hire Neighebour.
1477 Paston Lett. 794 III. 186 Putt hym away by
cause he is daungerous.
1454 Paston Lett. No. 223. I. 311 Cosetheryng
that youre doutyr is desendyd of hym be the modyr syde.
Loss of inflections / Rise
of verb phrases
Inflected
vs periphrastic subjunctive
Already
variation within OE, e.g. in subordinate noun clauses / ‘dependent desires’
§ ic wylle ţt he wunige
đus ođ ic cume
§ ...wiscton ţćt hi moston
swa wunian ođ ende
In
ME, periphrasis increases
o
“The gradual
erosion of verbal inflections made it necessary to replace the subjunctive by
something more transparent.”
o
“The early use
of the periphrastic construction may be due to a desire to be more emphatic and
possibly to be more specific than was possible with the subjunctive form.”
Examples
in main clauses
God gyve ţet ure ende beo god”
c1430:
“now ţrift and ţedom mote ţou have my leve swete barn”
prosperity beloved
Examples
in dependent clauses “particularly after volitional expressions (wishes,
exhortations and commands”
o
inflected
§ God gyve ţet ure ende beo
god”
§ I wisshe ţanne it were
myne”
§ Loke ţat ţou wite wel
who do mikel or litel
o
periphrastic: should,
mote, would
§ “gladly hym biddes ţat his
hert and his honde schulde hardi be boţe” (GGK 371)
“and prayed her to han good fame And that she nolde
doon hem no shame”
Other
verb phrases:
How
is the ‘future’ marked in ME?
o
with the finite
non-past
although it be soure to suffre, ţere cometh
swete after
o
with shal
and wil (although these “remain modally marked”—conveying necessity or
desire – and aren’t grammaticalized)
shal
· (OE sculan ‘to have
to’): obligation, commands
o
very common in
the third person
· ‘In ME often used .in all
the three persons to indicate that an event is ordained to take place in
accordance with divine will or fate” (Mustanoja 491): e.g.
“And rightful folk shul
gon, after they dye, To hevene”
· -> use in prophecies and
predictions
o
develops
earlier as a future marker
o
not related to
the will of the subject
wil:
o
(OE willan
‘to want to’)
o
“since modally
it is connected to the desire of the speaker/subject, occurs far more often in
the first person”
o
“may have
developed out of its use in habitual contexts”
He is a fool that wol foryete hymselve
o
if the subject
is inanimate, it’s lost its modal meaning
And I, boke, wil be brent,
But Jhesus rise to lyve
o
Some examples
are clearly not ‘desire’!
Oure manciple, I hope he wil be deed (northern meaning of hope)
How
could ongoing action in the present be expressed in ME?
o
with the finite
non-past:
“What! Alison! Herestow nat
Absolon,
That chaunteth thus
under oure boures wal?”
o
in phrases with
BE + ing participle
“As Cancacee was pleyyng
in hir walk
There sat a faucon over hire
heed ful hye.”
Optional
in ME constructions where it’s not permissible in PDE:
o
habitual action
“Arestotill sais ţat ţe bees are feghtande
agaynes hym ţat wil drawe ţaire hony fra thaym”
o
imperative
“John, be thou here abydand”
o
stative verbs
“The tour ... was joynynge in the wal to a
foreyne” [adjoined ... privy]
Factors
promoting the development of the ‘progressive’ aspect
o
Loss of OE
prefixed verbs and the distinctions conveyed by no prefix
o
prefixed verbs
conveyed perfectivity: forbaernan ‘burn up’
o
non-prefixed
verbs conveyed duration: baernan ‘burn’
§ once the distinction was
lost, new strategy needed?
o
Conflation of
present participle and verbal noun
o
OE He is on
huntunge, He is huntende ->
o
ME He is
hunting
Other
issues, not to be covered in class
“That” and “wh-“ as relative
pronouns
How
did that become a relative pronoun?
It
already was in OE, in its role as a demonstrative pronoun (cf. se, seo,
etc.)
Stephanus
is Grecisc nama, ţćt is on Leden Coronatus, ţćt we cweđađ on
Englisc, Gewuldorbeagod
That in ME
§ with the loss of case forms,
that emerged as the demonstrative
§ se got levelled to the
and took over the function of the definite article
§ the old relative
subordinator ţe disappeared. Reasons offered include
o
phonologically
rather weak
o
confusion with
definite article
o
that
was well entrenched as a conjunction in other sorts of clauses (e.g. purpose
and result)
Indeed,
in ME, that was the ‘general subordinator’: added to all kinds of
conjunctions
§ now that, if that, when that
§ before that, in that
How
did interrogative pronouns become relative pronouns? Probably via
1.
Use of interrogative pronouns in indirect questions:
ţa ascode he hwa ţćr ferde. ‘then asked he who there went’
2.
Pronoun loses interrogative character and becomes a ‘generalizing’ relative
(‘whoever, whichever’; cf OE swa hwćt swa ‘whatsoever’):
hwam mai he luue treweliche hwa
ne luues his broţer
whom may he love truly, [he] who
does not love his brother
3.
Pronoun has a clearly definable antecedent:
for Adames gulte, to hwam ure
Drihten seide...