^{Note 1 in page 377} Harry R. Warfel, “Syntax Makes Literature,” CE, xxi (1960), 251–255.

^{Note 2 in page 377} Francis L. Utley, “Structural Linguistics and the Literary Critic,” JAAC, xviii (1960), 317–328.

^{Note 3 in page 377} Roman Jakobson, “Linguistics and Poetics,” Style in Language, ed. Thomas A. Sebeok (Cambridge, Mass., 1960), p. 375.

^{Note 4 in page 377} The Language Poets Use (London, 1962), p. 10.

^{Note 5 in page 377} But see Donald Davie's Articulate Energy (London, 1955), a theoretical and somewhat polemical discussion of the role of syntax, pseudo-syntax, and non-syntax “articulation … through images spaced by rhythms” (p. 161), and Francis Berry's historical survey of certain grammatical usages in Poets' Grammar (London, 1958). Mrs. Nowottny's book, as its title indicates, covers all aspects of poetic language.

^{Note 6 in page 377} Walter Jackson Bate, The Stylistic Development of Keats (London, 1959; first pub. 1945), pp. 96–97.

^{Note 7 in page 377} E.g., his Linguistics and Literary History: Essays in Stylistics (Princeton, N. J., 1948), and “Language of Poetry,” in Language: An Enquiry into Its Meaning and Function, ed. Ruth Nanda Anshen (New York, 1957). For a discussion of the strengths and shortcomings of the “New Stylistics,” see David Lodge, Language of Fiction (London and New York, 1966), pp. 52–56.

^{Note 8 in page 377} E.g., The Verbal Icon (Lexington, Ky., 1954).

^{Note 9 in page 378} Stephen Ullmann, Language and Style (Oxford, 1964); Randolph Quirk, The Use of English (London, 1962), pp. 239 ff., and “Poetic Language and Old English Metre,” in Early English and Norse Studies, eds. A. Brown and P. Foote (London, 1963). An intensive critical application of syntactic analysis is the splendid essay by Allan Rodway, “By Algebra to Augustanism,” in Essays on Style and Language, ed. Roger Fowler (London, 1966), pp. 53–67; see also my “Syntactical Analysis and Old English Poetry,” NM, lxiv (1963), 373–378.

^{Note 10 in page 378} For statements to this effect, see, e.g., Richard Ohmann, “Generative Grammars and the Concept of Literary Style,” Word, xx (1964), 423–439, and M. A. K. Halliday, “The Linguistic Study of Literary Texts,” Proceedings of the Ninth International Congress of Linguists (The Hague, 1964), pp. 302–307.

^{Note 11 in page 378} Since my interest here is only in the application of grammatical description to poetry, I ignore such essays as S. Chatman's “Robert Frost's ‘Mowing’: An Inquiry into Prosodie Structure,” KR, xviii (1956), 421–438.

^{Note 12 in page 378} “Linguistic Theory and the Study of Literature,” in Essays on Style and Language, pp. 1–28, and “Linguistics, Stylistics: Criticism?” Lingua, xvi (1966), 153–165.

^{Note 13 in page 378} The term is associated with Halliday (see n. 10). As Fowler observes (Lingua, p. 163), stylistics “is a tortured subject. The definition of ‘style’ is of course a sitting target for the linguist, and too often he progresses no further.” In his n. 19, Fowler lists some “notable tormentors.”

^{Note 14 in page 378} Fowler, Lingua, pp. 155–156.

^{Note 15 in page 378} “Linguistic Theory …,” p. 27.

^{Note 16 in page 378} Lingua, pp. 162–163.

^{Note 17 in page 379} A fine piece of criticism stemming from linguistic description is Geoffrey Leech's “ 'This bread I break'—Language and Interpretation,” REL, vi (1965), 66–75. Independently from Fowler, Leech follows the same “program,” being selective and purposeful in the items he describes, and proceeding from linguistics to stylistics to criticism. He incorporates the concept of foregrounding from the Czech School (see the articles by B. Havrânek and J. Mukarovsky in A Prague School Reader, trans. Paul L. Garvin, Washington, 1958), which he interprets as “motivated deviation from linguistic, or other socially accepted norms” (p. 68); and by treating the described “foregrounded aspects of a poem as so many question marks,” he proceeds to interpret Thomas' poem “cumulatively,” remarking that “by such detailed observations … it is possible to see a basis in linguistic observables for those most elusive of critical concepts: climax, resolution, artistic unity” (p. 74). It is instructive to compare the methodology in Leech's essay with that in A. A. Hill's earlier structural analysis: “An Analysis of The Windhover: An Experiment in Structural Method,” PMLA, lxx (1955), 968–978.

^{Note 18 in page 379} That the “how” can be separated from the “what” for linguistic description seems to be a cardinal principle of the new linguistics; see the Ohmann and Halliday articles cited in n. 10.

^{Note 19 in page 379} There appears to be a lack of clarity about “semantics”: Fowler and followers of Halliday apparently equate it with the study of the connection between context and particular linguistic forms, context “comprising all relevant things outside language” (Fowler, Lingua, p. 156)—a different thing from lexis, which involves, inter alia, a study of the frequency of collocations of lexical items, something of a statistical matter. Chomsky and others (e.g., Paul Ziff, Semantic Analysis, Ithaca, N. Y., 1960; J. Katz and P. Postal, An Integrated Theory of Linguistic Descriptions, Cambridge, Mass., 1964) equate semantics with the selection from dictionary meanings for a string of particular words—a sentence or an utterance—as determined by the syntax in which the words are embedded.

^{Note 20 in page 379} The Hague, 1962.

^{Note 21 in page 381} In a different context, Leech distinguishes between semantic parallels in such a statement as Absent in body, but present in spirit (i Corinthians), where absent and present, and body and spirit, show a “systemic semantic contrast,” and the parallelism in I kissed thee ere I killed thee, where kissed and killed “are in no easily definable semantic relationship; their antithetic reinforcement comes rather from emotive contrast and from dramatic context.” This distinction, Leech remarks in a footnote, “presupposes a structural approach to referential semantics which Levin, in common with many other linguists, does not entertain.” See “Linguistics and the Figures of Rhetoric,” in Fowler's Essays, pp. 154–155. To this might be added that the semantic equivalence Levin sees in he painted the house and whitewashed the garage (see text, above) is not so simple as he supposes; for if the and were changed to but, painted and whitewashed would become antonymous and the dissimilarities between house and garage would be foregrounded instead of their similarities. The meaning of the conjunction, which Levin ignores, is thus very important. For a more elaborate criticism of Levin, see Nicolas Ruwet, “L'Analyse structural de la poésie,” Linguistics, n (1963), 38–59, esp. 49 ff. Subsequent to the acceptance of this paper, in a review of Fowler's Essays in EIC, xvi (Oct. 1966), 458–459, Mrs. H. H. Vendler makes the exact same point as I have above about Levin's “false” transformation of All losses are restored, showing additionally the significance of this passive construction for the meaning of the whole sonnet.

^{Note 22 in page 382} In Fowler, Essays, pp. 68–81.

^{Note 23 in page 382} The question of the relation of poetic grammar to “normal” grammar is an interesting one. See S. R. Levin, “Poetry and Grammaticalness,” in Proceedings of the Ninth International Congress of Linguists (The Hague, 1964), pp. 308–314;; j. P. Thome, “Stylistics and Generative Grammars,” Journal of Linguistics, i (1965), 49–59; G. N. Leech, in Fowler's Essays, esp. pp. 136–144.

^{Note 24 in page 383} I would not claim that all subject/object displacements are significant in this way, but context can make them so, more often I suspect than has been realized; cf. my remarks on the switch from / to me in a sentence of the OE Wanderer, in my NM article (see n. 9). Mrs. Vendler, in the review referred to in n. 21, says that Larkin's poem is “particularly inept” and “bathetic” (p. 460), and censures Sinclair for even wasting his time on it. But her evaluation of the poem is based on conceptual and contextual grounds, and even a poem judged banal for such reasons may be interesting linguistically. This only suggests that “evaluation” is not part of the linguistic prerogative, and no linguist would, I think, claim otherwise. And though I have in this paper criticized Levin and Sinclair, I nevertheless think both deserve the critics' gratitude for trying to establish (however horrendous to critics' tingling ears their terminology may be) the formal bases of a poem's linguistic aesthetic.

^{Note 25 in page 383} A good example of the interaction of word order and meaning appears in Rodway's article in the Fowler collection where, analyzing a couplet of Crabbe's, “Who far from civil arts and social fly / And scowl at strangers with suspicious eye,” Rodway suggests that though the placing of the verb at the end of the first line “has a purely aesthetic effect, like a variation in a tune,” the placing of scowl near the beginning of the second “elegantly illustrates (without vulgarly imitating) Crabbe's turning from the villagers' negative, to their positive hostility” (p. 59).

^{Note 26 in page 384} The earlier versions reproduced below are taken from Giorgio Melchiori, “Leda and the Swan: The Genesis of Yeats' Poem,” EM, vii (1956), 148–150, who gets them from Richard Ellmann, The Identity of Yeats (London, 1954), pp. 176–179. Melchiori does not discuss at all the significance of the grammatical changes; the best criticism of the sonnet is undoubtedly Leo Spitzer's, “On Yeats's Poem ‘Leda and the Swan’,” MP, Li (1954), 271–276, which does incorporate syntactic analysis, though it makes no reference to the earlier versions. But the points I have made above, and make below, are mostly my own.

^{Note 27 in page 384} “Linguistic Study,” pp. 305–306.

^{Note 28 in page 384} John Spencer, “A Note on the Steady Monologuy of the Interiors,” REL, vi (1965), 32–41.

^{Note 29 in page 384} Old English poetry, with its oral-formulas and other highly developed stylistic conventions, poses special problems for the critic; but these are not so limiting as many Old English scholars suggest they are. Cf. Frederic G. Cassidy, “How Free Was the Anglo-Saxon Scop?” Franciplegius, eds. J. B. Bessinger and R. P. Creed (New York, 1965), pp. 75–85.

^{Note 30 in page 385} The above paragraph is taken almost verbatim from my essay “Grendel's Approach to Heorot: Syntax and Poetry,” to appear in R. P. Creed's volume, New Approaches to Old English Poetry.

^{Note 31 in page 385} Postscript on Beowulf (Cambridge, Eng., 1948), pp. 74–75.

^{Note 32 in page 386} I am not suggesting that these different meanings are “fixed” for this syntactic pattern, either in OE or NE. It takes only a glance at the OE Judith or Maldon, for example, to see that their poets did not use the same patterns in similar narrative circumstances. But the meanings inhere in the patterns, and they can be foregrounded; and it would seem that the Beowulf-poet felt them there, and so foregrounded them.

^{Note 33 in page 386} Cf. this comment from J. Spencer and M. J. Gregory, in Linguistics and Style (London, 1964), p. 94: “Lexical ambiguity has long been recognized in the metaphor, the pun, and other types of imagery, but it is important to observe that ambiguity is not restricted to lexis. The syntax of poetry probably deserves more attention than it has hitherto received [in this respect].” They offer no examples, however.

^{Note 34 in page 387} Cf. R. Quirk, The Use of English, pp. 249–250, on ordinary and poetic syntax in the opening of Eliot's Wasteland.

^{Note 35 in page 387} For some discussion of Malkin's and Blake's own revisions of the fourth line: “forged thy dread feet” and “formed thy dread feet,” revisions which support the potential subject-object interpretation of the line, see M. K. Nurmi, “Blake's Revisions of The Tyger,” PMLA, lxxi (1956), 678–679, n. 16, and J. E. Grant, “The Art and Argument of ‘The Tyger’,” in Discussions of William Blake, ed. J. E. Grant (Boston, 1961), p. 70. I had not seen the latter essay when I wrote the above, and I think it interesting that Mr. Grant, too, feels “that the indeterminate syntax intimately relates the creator and his creature,” though he has not analyzed the stanza in terms of linguistic expectation, as I have done.

^{Note 36 in page 387} Language of Fiction (see n. 7), p. 56.

^{Note 37 in page 387} This paper is a revision of a Special University Lecture delivered at the University of London in May 1966.1 wish to express here my deep appreciation to the John Simon Guggenheim Memorial Foundation for its financial support, and to Randolph Quirk for his time, patience, and tactful criticism.