Professional Documents
Culture Documents
GEORGE L. DILLON
1 This and subsequent citations of Quintilian from the Loeb edition, edited by H.E. Butler
(1959). Here 1X.1.13. Citations of Ad Herennium (Ad H.) also from Loeb edition, edited by
Harry Caplan (1954).
’ One of these, Scrambling (for Latin), Ross suggests relegating to a special ‘stylistic com-
ponent’ of the grammar rather than the normal transformational component (1967: section
3.1.2). In this he seems to be following Chomsky, who would also relegate Topicaliiation to
such a component (1965: 127-127,221-222). In Syntacfic structures, Chomsky suggests that
the products of Node-Raising should possibly be characterized as semi-grammatical (1957: 35,
note 2). None of this affects my argument in the least, since Ross has formulated the processes
1
2 C.L. Dillon/Literary transformations and poetic word order
there are two classes of grammatical schemes that delete and permute major con-
stituents, those described by Puttenham as working ‘by defect’ (Quintilian: detrac-
tio) - primarily the zeugma family - and those working ‘by disorder’ - hyper-
baton, anastrophe, hysterologia, cacosyntheton, and hysteron-proteron. Rubel
(1941) and Joseph (1947) have shown that the schemes catalogued by Puttenham
(among others) are quite characteristic of the poetic languages of Spenser and
Shakespeare. By analysing these schemes as transformations and comparing them
to usual optional transformations of English, we can describe precisely what is
‘poetical’ in the syntax of Spenser and Shakespeare (and those who imitated them)
and how it is related to ‘usual English’ of the Renaissance. Further, this analysis
establishes the non-self-evident point that the re-orderings and deletions described
by the rhetoricians are constrained and systematic in their operation. It is not the
case that even Spenserian word order is ‘free’, or that ‘poetic license’ enables one
to do anything he wants in poetry. That is, a transformational model of grammati-
cal schemes can explain how the literary language works.
(Note: I will draw on works in the rhetorical tradition from Quintilian and Ad
Herennium through Puttenham (1589), making use of the valuable works by Ian-
ham (1968), Rubel (1941), and Joseph (1947) to thread the maze of terminology.
I have adopted the policy of ‘second-guessing’ individual rhetoricians in a few cases,
either because their descriptions are imprecise or conflict with another authority,
but I have done so in favor of the examples: when the examples do not examplify,
I have substituted others that do. I have limited the present analysis to schemes
which re-order and delete major constituents (N, V, ADJ).)
2. ZEUGMA SCHEMES
Zeugma schemes are classified as schemes of defect or ‘taking away’ in the rheto-
rics along with ellipsis (omission of minor words such as the that complementizer3).
Zeugmas are reductions of conjoined sentences. I will describe them in transforma-
tional rather than Interpretive terms as derived by deletion and Conjunction Reduc-
tion, referring to the constituent that is not deleted (i.e., that ‘deletes the others’)
as the Controller. I assume some familiarity with the transformations mentioned,
as transformations and shown that they obey certain constraints on the operations of trans-
formations. Jacobs and Rosenbaum (1971) discuss WHCleft, Extraposition, even Relative
Clause Reduction as ‘stylistic’, virtually identifying ‘optional’ and stylistic transformations.
There is considerable uncertainty in transformational grammar on what is to be classified as
‘stylistic’, and on this modern grammarians might learn a few tricks from the rhetoricians,
though few are really consistent about it either, the outstanding exception being Quintilian.
3 See George Puttenham, 77~ arte of English poesie (1589: 136), Hereafter ‘Putt’. I have
silently modernized spelling and set off the examples as is customary in linguistic studies today.
Spelling is also modernized in the citations of Spenser and Shakespeare.
G.L. Dillon/Literary transformations and poetic word order 3
For Latin, the first two zeugma-types seem cases of normal re-orderings and dele-
tions. If following Joan Maling (1972) we assume Gapping works only to the right,
Quintilian’s first example reflects the ‘Node-Raising’ option of Conjunction Reduc-
4 G.L. Dillon/Literary transformations and poetic word order
(4)
autv’\
Np/s\ NPHSIVP
/W\v
Y ‘v
,p? ,pY
P ;
pudor u turpitudine revocaverit metus a revocaven’t
CR(N-R option)
Y
aut \
NP~s‘NP NPAS\VP
I /“\
pudor a turpitudine metus a periculo revocaverit
This transformation, while it exists for English (see below), will not create the ef-
fect of gapping to the left, since the V is not a final constituent - hence hypo-
zeugma producing ‘left-gapping’ in English is not possible with the usual transfor-
mations and the usual word order. If the ‘left-gapping’ zeugma and the others are
normal deletions for Latin, however, there should not be anything particularly
‘unusual’ about them, and indeed Quintilian denies that they are schemes in Latin
because they are so common (vulgaria, 1X.3.64). It would be in his spirit, however,
to conclude that ‘left-gapping’ is a ‘literary’ scheme in English. It would be a mis-
take, though, to postulate a ‘Hypozeugma transformation’ for English, ‘for it can be
explained as the outcome of a ‘literary’ re-ordering transformation (see following
section).
The remaining case is the mezozeugma type:
(17) Who doth sometimes counsel take, and sometimes tea (Pope, R of I,. 111.7-8).
(18) Faire maids’heauty(alack) with years it wears away,
And with weatherandsickness, and sorrow as they say (Putt. 37; the Left-Dislocation is
irrelevant here).
(19) Formae dignitas aut morbo deflorescit aut vetustate (= 16).
Assuming with Stockwell et al. (1972) that all conjunction derives from sentential
conjunction, these would be derived by Gapping and Conjunction Reduction (see
Appendix E). Alternatively, these might be derived by Conjunction Reduction and
G.L. Dillon/Literary transformations and poetic word order 5
then ‘Poetical-Conjunct Movement’ which would differ from the suggested rule of
Conjunct Movement in English (the existence of which is doubtful). This seems un-
necessarily wasteful, given the existence of Gapping. In short, given the derivation
in appendix E, no special schematic transformation need be postulated here.
The distinction drawn by Quintilian between ‘ordinary’ optional deletions and
true schemes appears in a different form in Puttenham in his discussion of hype-
.zezCxis,where he makes it clear that some deletions (or in this case pronominaliza-
tions) are normal - then non-deletion is ‘schematic’:
Unto the king she went, and to the king she said,
Mine own liege Lord, behold thy poor handmaid.
Here [went to the king] and [said to the king] be but one clause iterated with
words of sundry supply. Or as in these verses following:
Here [my Zadygave] and [my lady wist] be supplied with iteration “by virtue of
this figure” (Putt. 139). Presumably Puttenham wants the second my lady to be
pronominalized, as it is in the last line. Putter&am does seem to regard gapped sen-
tences right or left and sentences with split conjuncts as sufficiently out of the or-
dinary as to qualify them as schematic.
Latin is a verb-final language, Quintilian informs us, and any variation from that
is an hyperbaton and ‘schematic’ (1X.4.23). The ordering principles of the modern
vernaculars, however, are quite different, and the Renaissance rhetoricians show
considerable originality in finding native examples of what they understood their
classical sources to be describing. There is great variation in terminology in this
area: Rubel uses hysteron-proteron for the generic term, Joseph uses anastrophe,
apparently following Angel Day.4 But as Quintilian himself said in a slightly dif-
ferent connection, it does not matter what you call them - what matters is what
they do (IX.i.7-8).
Joseph reports that Angel Day cites the following as a type of anastrophe:
4 For Quintilian, anastrophe is a minor kind of hyperbaton involving the switching of two
words (a word-level metathesis called in Puttenham ‘hippelage or the changing’).
6 G.L. Dillon/Literary transformations and poetic word order
s
NP---
NPYS\VP
v/ ‘PP
I’\ I P
faults no man liveth without
(25) Whom suddenly he wakes with fearful fright (Spen. FQ. 15.4).
(26) Her all in white he c&d (FQ. I.i.45).
(27) Her soon he overtook, and bad to stay (FQ. I.ii.20).
(28) There is virtue in that Falstaft him keep xith, the rest banish (Shak. I Hen. IV. II.iv.
424).
(29) WhomZserve above is my master (A W. II.iii.26).
(30) Your party in converse, him you would sound,
He closes with you (Ham. II.i.42).
(3 1) My life’s foul deed, my life’s fair end shallfree it (Shak. R. of Lu. 1208).
(32) That thing you speak of Z took it for a man (Lear IV.vI.77).
(33) Rut if Z thrive, the gain of my attempt
The least of you shall share his part thereof (Ric. ZZZ. V.Iii.267).’
The peculiar perspective of the Renaissance Humanist is evident here in two ways:
he gives no consideration to French, the probable model for this construction, and
he does not recognise that the rule does not seem to have been obligatory in earlier
English, in fact, rather the other way around for adjectives of French and Latin
origin (see Traugott 1972: 158; also Abbott, Q 419 for many examples from Shake-
speare). This situation would be noted in generative grammar either by making ad-
jective-preposing optional, or marking Latinate and learned French words as excep-
tions to the rule. For Puttenham, however, non-application of ADJ-preposing is a
‘scheme’.
On the other hand, Puttenham also classifies as a vice the application of ADJ-
preposing out of a complex adjectival. Under the heading of hysteron-proteron
5 Note here that ‘Left-Dislocation’ does not move the NP ail the way to the front - hence it
is not strictly a ‘root’ or ‘post-cyclic’ operation here (see Emonds 1970: 18-19). Ross, how-
ever, has some doubts about restricting this T to last or post-cyclic operation (1967: 6.2.1).
Similar remarks apply for (28).
8 G.L. DillonfLiterary transformations and poetic word order
it may be done, either by a single word or by a clause of speech: by a single word thus:
For perform not: and this vice is something tolerable enough, but if the word carry any notable
sense, it is a vice not tolerable [note the postnominal ADJ - CLD] , as he that said praising a
woman for her red lips, thus:
Which is no good speech, because either he should have said no more but a coral lip, which had
been enough to declare the redness, or else he should have said, a lip of coral hue, and not II
coral lip of hue. Now if the disorder be in a whole clause which carrieth more sentence than a
word, it is then worst of all (Putt. 213).
(See Abbott, 8 419a for further examples.) Puttenham’s remark is valuable in iden-
tifying this construction as distinctively ‘schematic’ by his time, since Traugott cites
examples of it from writings thought to reflect informal usage (1972: 159). Putten-
ham’s remarks indicate that by his time, ADJPreposing rules were much like those
for NE (or at least were to him), however frequently they may have been violated.
Franz (1939: 8 685) notes that non-preposing in Shakespearean poetry often has a
metrical motive and that learned authors employ it more than more down-to-earth
writers in prose.
3.3 SO V order
One aspect of Renaissance literary syntax not remarked on by Putter&ram is the
fairly common SOV order in dependent clauses, especially adverbials and relatives. I
cite representative examples from Shakespeare and Spenser (and Pope!):
6 Note that sometimes the AUX ‘goes along with’ the V, sometimes it remains behind and
do-support is triggered, as in this line. Spenser does each. See Appendix F.
G.L. Dillon/Literary transformations and poetic word order 9
(For other examples, see Rubel 1941: 36, 44, 64-65, 163, 182, 245, noted as
Hysteron-proteron.) The construction has strong native roots, for it was the normal
order in dependent clauses in OE (as in Modern German) - as with postnominal ad-
jectives, an element of ‘archaism’ may be involved. The evidence is fairly clear, how-
ever, that even by Chaucer’s time this order was obsolete for prose (and presumably
for ‘common’ usage).’ In an interesting comment on Tottel’s revisions of Surrey’s
Sonnets, Rubel notes that SOV orders seemed to be coming into literary for the
Elizabethans: “he [Tottel] occasionally shifts a perfectly simple line to produce
slight hysferon proteron, so popular with later poets:
(Surrey): That Cupid’s scourge did make me run
(Tottel): That Cupid’s scourge me caused to run” (Rubel 1941: 64-65).
We can reasonably conclude that SOV order was possible (or that reordering to
SOV was a scheme) in Renaissance literary language in dependent clauses. Actually,
Spenser uses it heavily outside of dependent clauses as well, though in this he seems
somewhat unusual.8 In the first canto of the Fairie Queen I have come across these
examples (and the list is not exhaustive):
(41) But the other half did women’s shape retain (14)
(42) The stroke down from her head unto her shoulder glaunst (17)
(43) He fair the knight saluted (30)
(44) Her swollen heart her speech seemed to bereave (52)
If we assume that the basic constituent order of English is SVO, we need a rule like
that suggested by Ross (and after him, Bach 197 1) for dependent clauses in German
(‘Verb-Final’ - see Appendix) which re-orders 0 and V. In Modern German the rule
is fairly simple, moving the AUX t V to the end of the clause (i.e., across all of the
Comp). Certain examples from Shakespeare and Spenser suggest that the re-ordening
rule necessary for their English is a variant of Verb-Final which moves either V or
AUX + V either to the end of the clause or just to the right of the first NP (the
direct object):
’ Walarlan Swieczkowski (1962) finds that while Piers Plowman shows 37% V-final order in
dependent clauses, the roughly contemporaneous prose sermons in Middle English sermons have
virtually none. A better indicator of ‘common’ usage would be the Paston Letters.
’ Though Swieczkowski notes a significant percentage (24%) of verb-final main clauses in
Piers Plowman (1962: 53). Hence Spenser’s use of this order in main clauses may be an archa-
ism .
10 G.L. Dillon/Literary transformations and poetic word order
(47) Yet he her followed still with courage keen (FQ. III.iv.Sl)
(48) He piped grace, whilst they him doncst about (Mut. vii.46)
However, there are many cases in Spenser where ‘Verb-Final’ does seem to be
correct:
(49) The fiends do quake, when any him to them does name (FQ. III.iii.12)
(50) 7’he stroke down from her head unto her shoulder glaunst (= 42).
Apparently the rule should state that the V moves to the right over a variable which
must begin with NP.
On the other hand, (48) suggests that the movement should be of the NP rather
than the Verb, since a preposition is apparently stranded. It is impossible, however,
to move the entire complement by a transformation, since it is not a constituent.
And (48) is by no means unique :
(51) For fear lest &y should look their shows upon (Shak. MND. III.ii.385).
(52) For my good will is to k
And yours it is against (Temp. III. i. 31).
These examples of stranded prepositions are not decisive evidence against deriving
the sentences by Variable-Verb-Final, however, because some non-monosyllabic
With as much speed as I could, I ran after the thief that had undone me
Another
He ran against with fury rage, the doors most strong,
here likewise the preposition against is joined to the verb, which according to the order of con-
struction, ought to have served to the noun, thus,
He ran with furious rage, against the doors most strong” (fol. iiii recta).
G.L. DillonfLiterary transformations ond poetic word order 11
(53) That path they take, that beaten seemed most bare,
And like to lead the labyrinth about; (FQ. I.i.11).
Given possible NPP order, these examples are not counter-examples to Variable-
Verb-Final but further evidence for it.”
Now, given the availability of SOV order, we can largely account for the exam-
ples of ‘left-gapping’ of which Spenser is also fond. The ‘Node-Raising’ option of
Conjunction Reduction also applies in English to final constituents, giving us
(54) Mary must kill, and John must clean, the game.
Mary is now, and Joan will be, asleep.
lo Traugott (1972: 109) notes that pronominal objects tend to precede the V in OE
whether in main or dependent clauses and notes that this tendency continues in ME and into
ENE poetry (1972: 160-166). Possibly the clear case-marking on pronouns encouraged the use
of Verb-Final. Her example from I Hen. IV has been listed here (= 28) as a case of Topicaliza-
tion - since the sentence is imperative and lacks a subject, the only difference between the
analyses would be the strength of the intonation break between him (and rhe rest) and the
verbs.
Variable-Verb-Final may be involved in the derivation of the 18% of the sentences in Piers
Plowman which Swieczkowski lists as “verb-medial” (1962: 54ff.).
12 G.L. Dillon/Literary transformations and poetic word order
with a sharp constituent break before the raised verb. This would apply only if the
V has been moved all the way to the end of the clause, thus predicting that such
things as
(55) Yet arms till that time did he never wield (FQ. I.i.1).
(56) On horseback used Trtamond to fight
And Priamond on foot had more delight,
But horse and foot knew Diamond to wield:
With curtax used Diamond to smite,
And Triamond to handle spear and shield
But spear and curtax both usd Priamond in field
(IV.ii.42 - the spelling of usd indicates that it is monosyllabic).
This order was much more common in ME: Swieczkowski finds that two thirds of
the sentences which have direct objects first have the order OVS in Piers Plowman
(p. 21) and 40% of those with direct object first in the prose Middle English
Sermons (p. 43). Note that lines in the two examples from Spenser would be
metrically irregular were it not for the major constituent breaks between time and
did, foot and knew: time would otherwise be a stress-maximum in a Weak position,
and the ‘heavy foot’ of knew Diamond frequently follows a major break (see Halle
and Keyser 1971: 168ff.).
G.L. Dillon/Literary transformations and poetic word order I3
(For some discussion of this surprisingly obscure transformation in NE, see Aissen
and Hankamer 1972). A preposed time or place adverbial seems to trigger Subject/
Aux inversion frequently in Shakespeare, even in prose, especially when the subject
is pronominal :
It may seem that with Topicalization, Subj/Aux inversion, and SOV re-ordering,
all orders of major constituents are possible, and that the notion that schematic
transformations express an hypothesis about possible orderings therefore collapses
into random poetic licenses. This is not true: at least two simple orderings are still
ruled out for English (VSO, VOS; VSO is of course a possible order in questions in
ENE, though according to Franz already somewhat elevated - 6 706a). Further,
certain gappings are ruled out. We can have
SO + SOV (= 8)
sv + svo (= 54)
svo + so (= 13-15)
SOV + SO (Reason to my Passion yielded, Passion unto my Rage, Rage to a
hastie revenge - Sidney, cited in Rubel 1941: 126)
but not
in the poets can be derived by one ‘literary’ re-ordering transformation and the in-
dependently needed rules of Topicalization, Gapping, and Conjunction Reduction
(including Node-Raising). Since these rules restrict the set of possible random order-
ings, they show that even Spenserian syntax is not the product of random poetic li-
cense, but in fact is quite similar, formally, to normal English, requiring about the
same amount of syntactic adjustment from the modern reader that Chaucer does
phonologically.
I would like to conclude with one or two reflections on the general literary func-
tion of these schemes, since they do much more than simply ‘ornament’ the poetry.
Both Topicalization and Variable Verb-Final have obvious ‘presentational’ or focus-
sing effect, foregrounding certain material that would otherwise appear later. They
also facilitate rhyme: this is very clear in Spenser, whose rhyme scheme strains the
resources ‘of English severly - the great majority of Verb-Final shiits get the verb
to the end of one line where it both rhymes and matches clause boundary with line
boundary, as in these lines:
which, without preposing of the PP, would have stress maxima in two W positions:
The point of all this is that it is not a curious historical accident that some of the
great English poets used these ‘literary transformations’ because they happened to
live in the Renaissance and the Renaissance happened to be interested in classical
rhetoric. Further, since their usefulness is clear, and the transformations involved
occur in many languages, we would expect them to occur in other languages. In this
regard a remark in the DudenCrammatik is intriguing: “Endstellung der Personal-
form in Hauptdtzen war in Althochdeutschen hgutig. Sie war vor allem in der
klassischen Dichtung als Stilmittel noch beliebt. Heute ist sie veraltet: . . . das
Miidchen kam und nicht in acht das Veilchen nahm (Goethe). Und hinein mit be-
dachtigem Schritt ein Lowe tritt (Schiller)” (0 7010). It would seem that the kind
of analysis made here for Elizabethan literary syntax could profitably be carried
out in a number of languages.
I would like to conclude with a note of qualification. It is not the case that all
poets who used these schemes observed the same constraints on their use that apply
in modem English. Milton, in particular, applies Topicalization to one member of a
conjoined 0 and to the Complement of a Noun:
In normal usage, these are Islands exempt from ‘chopping’ (see Ross 1967, esp.
p. 111; Stockwell et al. 1973: 463f. for discussion). Milton also allows Variable-
Verb-Final to split conjoined OS and N + Comp:
For we to him indeed a11praises owe
And daily thanks (PL. IV.444-445)
Yet another is this moving of relative clauses arbitrarily far away from the heads
they modify
and to apply Relative Clause Reduction (IVH + be deletion) when so removed (this
is also impossible in modern English):
APPENDIX ON TRANSFORMATIONS
A. Conjunction Reduction
CR is actually a rule-schema. I wilI follow the formulation of Stockwell et al. (1972) which
includes Node-Raising as one option, but will show that option under B.
Very informally, CR properly analyses a conjoined pair (or triplet, etc.) of sentences into
corresponding B and C parts, then Chomsky-adjoins each B and C part:
(2)
and vs\ - and //SK
Np/s:p &+
w A SA
I&P ii-P Lp iJP
I I I I
hours pass away days pass away hours days pass away pass away
B C B C
G.L. Dillon/Literary transformations and poetic word order 11
Node-relabelBng relables co-ordinate constituents of the same type, ‘moving the label up’, and
Identical Conjunct Deletion ‘CoBapses identical conjuncts:
Y ‘i” ‘i” ;” B I
hours days pass away pass away hours days pass away
aut formae dignitas a morbo deflorescit aut formae dignitas a vetustate deflorescit
CR
-
formae dignitas aut a morbo aut a vetustate deflorescit
TOP
-
out morbo aut vetustate formae dignitas deflorescit (~1)
B. Node-Raising
The CR scheme does not require B to be a constituent. When C of the proper analysis is
a final constituent only, the effect is of ‘Node-Raising’ (Maling, 1972 - Ross’ ‘Conjunction
Reduction’):
This option derives (4), as noted in the text. If this option is taken for
au t formae dignitas a morbo deflorescit aut formae dignitas a vetustate deflorescit
B C B C
18 G.L. Dillon/Literary transformations and poetic word order
bP
formae dig.
I\
a morbo formae dig.
T
a vetustate heflorescit
C. Scrambling
Ross’ Scrambling, which applies to Latin but not to English, reorders the constituents of
major nodes (1967: 3.1.2). Thus for Latin, assuming SOV order, we can get SVO, OVS, VOS.
Scrambling applied to the previous structure would re-order V and S, giving
In Latin literature, the kind of radical re-ordering of constituents cited by Ross as motivation
for Scrambling is distinctive of poetry. The rhetoricians all note it, particularly the effect of
separating ADJ and N modified, usually citing Virgil (Quin. VILvi.62-67; Ad Herennium
IV.44). Quintilian cites Aeneid 1.109 as an example of excessive or confusing “trangressio”.
This criticism appears in Peacham (1577) as well, under the heading of synchisis, where Aeneia
I.195 is cited as an example of confusing re-orderings (G i recta).
D. Gapping
Following Maling (1972), I assume Gapping works only left-to-right in conjoined structures
to delete the second occurrence of a V (and third, etc.):
(13)
zs>s,
NP
7,
‘; “1”
her beauty pierce mine eyes her speech (Gapped) mine heart
G.L. Dillon/Literary transformations and poetic word order 19
libido pudorem vicit & audacia timorem vicit & amentia rationem vicit
vicit pudorem libido & vicit timorem audacia & vicit rationem amentia
E. ‘Split conjuncts’
The ‘split conjuncts’ of (16-18) seem to be derived by Gapping and CR. For (16)
auty ‘1
Np/s\
NPAS’vP
Y\
PP’ \v
K ‘i A I
formae dig. a morbo dejlorescit formae dig. a vetustate dejlorescit
T<HS\
s’ 1s s~------b
IbP IkP +P
I
formae dig.
I formae dig.
Piw-
a morbo
v
dejiorescit a’ vetustate
Note that (18) could be derived with SVO order, then Verb-final. This produces the same
derived structure, but simply shows that the input order is irrelevant to the derivation of these
sentences (though it is relevant to ‘left-gapped’ types (6, g-10)).
F. Variable verb-final
In Elizabethan poetry, this rule is variable in two ways: (1) over the amount of Comple-
ment moved over by the V. This can be formulated as
VY
1 2 =*where Y = [NP . . .] and does not include # (clause boundary)
02+1
20 G.L. DillonlLirerary transformations and poetic word order
(2) whether AUX ‘goes along with’ the V or not. Combining these:
XAUXVY
1 2 34=
I 2/Q 9 4+@2 +3
i. NPAUX NP v z (= 45)
ii. NP NP AUX v z (= 46,47)
iii. NP NP Z AUX V (= 48-50)
iv. NP AUX NP Z V (She had this knight from far compelled - FQ I.i.5) (= 8,9)
<I am assuming Y’ here is a cover-term for V, ADJ (predicate) or participial). Notice that this
is unlike the Germanic Verb-Final rule mentioned in Bach and Ross in that the AUX precedes
the V when they both shift. That is, this rule provides no evidence for ‘AUXs as Main Verbs’.
Finally, if an ADV occurs between AUX and V, and AUX is moved, the ADV moves also
(cf. FQ. I.i.t-63). The rule should thus be revised to read:
REFERENCES
Abbott, E.A. 1970. A Shakespearian grammar. Reprint. New York: Dover Publications. First
edition, 1966.
A&en, Judith, and Jorge Hankamer. 1972. Shifty subjects: a conspiracy in syntax? Linguistic
inquiry 3 : 501-504.
Bach, Emmon. 1971. Questions. Linguistic inquiry 2: 153-166.
Bloomtield, Morton. 1963. A grammatical approach to personification allegory. Modem philol-
ogy 40: 161-171.
Chomsky, Noam. 1957. Syntactic structures. The Hague: Mouton.
Chomsky, Noam. 1965. Aspects of the theory of syntax. Cambridge, Mass.: M.I.T. Press.
‘Corniticius’. Rhetorica ad Herennium. Harry Caplan, ed. Loeb edition, Cambridge, Mass.: Har-
vard University Press, 1954. (=Ad H.)
DudenGrammatik. 1966. Paul Grebe, ed. 2nd ed. Mannheim: Dudenverlag.
Emonds, Joseph. 1970. Root and structure-preserving transformations. Ph.D. dissertation. Cam-
bridge, Mass: M.I.T. Available from the Indiana University Linguistics Club.
Franz, Wilhelm. 1939. Die Sprache Shakespeares. 4. Aufl. Haile and SaaIe: Niemeyer.
HaIle, Morris, and S.J. Keyser. 1971. English stress: its form, its growth, and its role in verse.
New York: Harper and Row.
Jacobs, Roderick, and Peter Rosenbaum. 1971. Transformations, style, and meaning. Waltham,
Mass.: Xerox.
Joseph, Sister Miriam. 1947. Shakespeare’s use of the arts of language. New York: Columbia
University Press.
Lanham, Richard. 1968. A handlist of rhetorical terms. Berkeley and Los Angeles: University
of California Press.
Maling, Joan M. 1972. On ‘Gapping and the order of constituents’. Linquistic inquiry 3: lOl-
108.
Peacham, Henry. 1577. The garden of eloquence. Fats. reproduction with introduction by
William G. Crane. Gainesville: University of Florida Press, 1954.
Postal, Paul. 1971. Cross-over phenomena. New York: Holt, Rinehart, and Winston.
Puttenham, George. 1589. The arte of English poesie. Fats. ed. 1971. New York: Da Capo
Press; Amsterdam: Theatrum Orbis Terrarum. (= Putt.)
Quintilian. Institutio oratoria. H.E. Butler, ed. Loeb edition. Cambridge, Mass. Harvard Univer-
sity Press. (= Quin.)
ROSS,John Robert. 1967. Constraints on variables in syntax. Available from Indiana University
Linguistics Club.
Ross, John Robert. 1970. Gapping and the order of constituents. In: Bierwisch and Heidelph
(eds.), Progress in Iinquistics. The Hague: Mouton.
Rubel, Verk L. 1941. Poetic diction in the English renaissance. New York: Modern Language
Association. Reprint 1966, Kraus Reprint Co.
Stockwell, Robert, Paul Schachter, and Barbara Hall Partee. 1972. The major syntactic struc-
tures of English. New York: Holt, Rinehart, and Winston.
Swieczkowski, Walerian. 1962. Word order patterning in Middle English. The Hague: Mouton.
Thomas, Owen, 1968. Metaphor and related subjects. New York: Random House.
Traugott, Elizabeth Closs. 1972. A history of English syntax. New York: Holt, Rinehart, and
Winston.
Valesio, Paolo. 1970. Hysteron proteron and the structure of this course.
Weinreich, Uriel. 1966. Explorations in semantic theory. In: T.A. Sebeok (ed.), Current trends
in linquistics. The Hague: Mouton.
22 G.L. Dillon/Literary transformations and poetic word order