What are these symbols? Click to learn more...

The Chomsky of Grub Street: Edward Bysshe and the Triumph of Classroom Metrics

Peter L. Groves
Department of English,
Monash University, Melbourne.
Email: Peter L. Groves

Copyright © Peter L. Groves 1999
Received: 29 March 1999; Published: 5 May 1999

Abstract: For three centuries Edward Bysshe's highly influential The Art of English Poetry (1702) with its notorious attack on traditional foot-based metrics has been dismissed by critics and literary historians as a piece of mindless hackwork, unthinkingly plagiarized from a French original: interestingly, this reputation was in the first instance the creation of Charles Gildon, the writer who more than anybody laid the foundations for the elevation of foot-based metrics into the item of standard mental furniture that it later became. So entrenched has this reputation become that (judging from their contradictory and wildly inaccurate accounts of it) his critics seem content to dismiss Bysshe's theory unheard. In fact it represents a serious and innovative attempt in the tradition of late C17 Royal Society empiricism to codify, for the practical purposes of the versifier, the metricality of the English heroic line, with particular reference to the practices of neoclassical poetry. In attempting, moreover, an explicit account of the conditions governing metrical form in terms of syntactic and phonological structure (given the rudimentary understanding of these matters then available), his system represents a remarkable (though sketchy and incomplete) prefiguring of modern linguistic or "generative" metrics. Given its demonstrable superiority as a descriptive tool over the gimcrack humanist alternative which has since become the orthodox form of scansion, the paper goes on to consider some of the reasons for the latter's ultimate triumph, including what might be termed the ceremonial function of traditional metrical description.

Keywords: Bysshe, Gildon, Art of English Poetry, traditional metrics, Lancelot, Quatre traitéz de poësies, metricality, generative metrics, ceremonial function.

The Chomsky of Grub Street: Edward Bysshe and the Triumph of Classroom Metrics

[H]owever minute the employment may appear, and whatever ridicule may be incurred by a solemn deliberation upon accents and pauses, it is certain that without this petty knowledge no man can be a poet. (Samuel Johnson, Rambler 88)


Edward Bysshe must have one of the least enviable reputations in English letters: that of the cloth-eared hack who by means of his often-reprinted Art of English Poetry (1702) succeeded in foisting onto the eighteenth-century reading public an "almost worthless" account of English metre (Sampson 499) which "prescribes all the wrong things and proscribes all the right ones" (Saintsbury History 2:192n.), simultaneously "unoriginal" and "perverse," and "less significant in the advance than in the retardation of metrical analysis" (Culler 878, 885). Bysshe's biggest offence seems to have been his notorious heresy that "[t]he Structure of our Verses, whether Blank, or in Rhyme, consists in a certain Number of Syllables; and not in Feet compos'd of long and short Syllables, as the Verses of the Greeks and Romans" (1). It was his commercial competitor Charles Gildon who first sprang to the defence of the calumniated iamb in his combatively-titled The Complete Art of Poetry (1718), fulminating for several pages against "the shameful ignorance of a little Pretender" (92) with his "gross Absurdities and visible Contradictions" (91).

Bysshe's crimes were compounded by the curious fact that his "shallow and indigested notions" (Gildon 93) remained in great demand throughout the century, becoming "very likely the most influential prosodic handbook ever written" (Brogan 243).<1> He and Gildon between them so dominated the market in vernacular gradus that their material was frequently recycled throughout the century. John Newbery (or rather his anonymous hack), for example, in his Art of Poetry on a New Plan, plagiarises both, successively; like a true hack he doesn't notice (or doesn't care) that the two kinds of metrical discourse are completely incompatible.

And yet while the modern jury is unanimous in its verdict on Bysshe, it seems oddly confused about the charges: for some he is a totalitarian pedant demanding "undeviating regularity of accentuation" (Omond 32) and "absolute regularity of iambic structure" (Culler 883), yet for others he is a mere hack reckoning syllables on his fingers, "so little aware of the structural importance of accent in English poetry that he felt it not worth wasting time on" (Fussell 12). Clearly at least one of these contradictory views must represent a strikingly inattentive reading of what is not, after all, a very long or difficult treatise. In fact both misrepresent it so recklessly that the phenomenon calls for explanation. Obviously it is Bysshe himself who is not worth wasting time on: since the critics and literary historians "know" his work to be intellectually nugatory, and only mention it because of its unaccountable popularity, they do not feel it necessary to give it any serious consideration.. This damning reputation is almost as old as the work itself: it was firmly established as early as the 1730s, when Hogarth could use the book as a signifier of the Grub Street hack (an open Bysshe sits at the elbow of his Distressed Poet) and to some extent it seems to have been the creation of Gildon's "venal quill."<2> It was Gildon who first suggested that the Art was mere plagiarised hackwork, "mostly borrowed from the Messrs. of the Port-Royal on the French Versification" (93) and "applied very ignorantly and falsely to that of the English" (293). More recently Culler has elaborated this charge (877):

from the French treatise, the "Breve Instruction sur les Regles de la Poësie Françoise" [Lancelot 49-77], he took almost his whole prosodical system. Indeed, his "Rules" are simply a translation and adaptation of the "Breve Instruction" with English examples replacing the French. . . . Thus Lancelot says, "Il n'y a point de Langue qui joigne si souvent plusieurs voyelles ensemble pour en faire des diphthongues, que la nostre," [59] which Bysshe echoes, "There is no Language whatsoever, that so often joyns several Vowels together to make Dipthongs of them, as ours." [13]

The picture meant to emerge is that of a hack like Newbery's, mechanically reproducing his source with no thought for its relevance or applicability to its new context. It is not, however, supported by the instance that Culler so triumphantly flourishes. Bysshe makes a defensible and relevant claim about English, which is particularly rich in diphthongs (the examples he cites are phonetically accurate), whereas Lancelot, writing of "[d]outes sur le nombre de syllabes de certains mots" [59], is at best irrelevant (he confuses orthography and pronunciation, citing "eau en beau," for example, as composed of three vowels) and at worst nonsensical (French has very few diphthongs). Though technically he has plagiarized Lancelot's turn of phrase, in effect Bysshe has turned base metal into gold.

Culler's mistake should give us pause, for while Bysshe might still be an ignorant hack who stumbled by chance on the truth here, the case remains unproven. At the very least we might, rather than simply re-cycling the derogatory labels placed on him by his commercial rival, want to reinvestigate his work on its own merits. We might even explore an alternative model, and attempt to read as a deliberate innovation his explicit rejection of the half-baked conventional humanist account of English versification that his age had inherited from the Elizabethans. His surprising use of French versification as a model may be considered the deliberate choice of a late-seventeenth-century empiricist, committed to observation rather than to reverence for tradition and auctoritas. After all, no contemporary who considered English versification with the sceptical mind-set of a Royal-Society man could help noticing that classical Latin provides a poor model for the description of vernacular metres: as Bysshe observes, "each Tongue has its peculiar Beauties, and . . . what is agreeable and natural to one, is very often disagreeable, nay inconsistent with another." (2). Central to classical Latin metre is the phonological feature of syllabic weight or "length," which functions as a kind of exchange-value, two so-called "short" syllables being in most places metrically equivalent to one "long" one, so that lines with the same underlying structure may have very different numbers of syllables. Literary English verse of Bysshe's period, on the other hand, resembles the modern Romance vernaculars that Lancelot treats of in being metrically isosyllabic: once rule-governed prosodic contractions and so-called "feminine" endings have been allowed for, all the lines in a given metrical tradition such as heroic verse will indeed have "a certain Number of Syllables."<3> There are other ways in which the vernacular metres are distinguished as a group from Latin quantitative forms: whereas quantitative metre governs the value of every syllable in the line, so that a single change of quantity (e.g. cano to plango in the first line of the Aeneid) will render the whole line unmetrical, vernacular metres are more tolerant, and specify syllabic value only at certain points. Similarly, while both word-stress and syntactic juncture play some role in all the vernacular metres, they are both irrelevant to the metrical structure of quantitative verse, which can "see" only syllabic quantity and word-breaks.

But while Lancelot's "Breve instruction" was a suitable place to begin, it provided no more than a starting-point for Bysshe: it is a serious misrepresentation to call the "Rules" "simply a translation and adaptation." The clearest evidence of this is the fact that the key structural elements of his theory, the notions of "Accent" and "Prevailing Accent," have no counterpart whatever in the "Breve instruction." For Lancelot "Le nombre des syllabes est donc ce qui fait toute la structure de nos vers"; (51)<4> Bysshe, by contrast, explicitly states that "'tis not enough that Verses have their just Number of Syllables: the true Harmony of them depends on a due Observation of the Accent and Pause" (4).

Bysshe's reputation as a mindless hack will need, then, at least some revision. I wish to argue in what follows that The Art of English Poetry represents a serious and innovative attempt to codify, for the practical purposes of the versifier, the metricality of the English heroic line, with particular reference to the practices of neoclassical poetry. In attempting, moreover, an explicit account of the conditions governing metrical form in terms of syntactic and phonological structure, his system represents a remarkable prefiguring of recent developments in the linguistic analysis of metre.


Bysshe's gradus ad Parnassum sold because it addressed a genuine need. The ability to dash off a few verses was not only a useful social accomplishment for the eighteenth-century gentleman, but also a marketable skill in an age with a seemingly insatiable appetite for exotic eclogues and poems on affairs of state. Not everyone with the ambition to turn a Persian tale for half a crown, however, had what Pope called the knack of versifying: Samuel Johnson "used once to be sadly plagued with a man who wrote verses, but who literally had no other notion of a verse, but that it consisted of ten syllables. Lay your knife and your fork, across your plate, was to him a verse" (Boswell 2:51). The would-be poet who turned to Gildon's Complete Art for instruction in verse-making, however, would have found only the "general Rule, that an Heroic verse in English, should consist of five Iambics", with the useless rider that "Mr. Dryden has frequently deviated with wonderful Beauty from this Rule, and by that gain'd a very considerable Advantage to his Versification" (300). While Gildon gives examples of variant feet -- trochees, pyrrhics, spondees, and so on -- he nowhere gives any rule (or even advice) about which "deviations" destroy the metre and which preserve it: what you are to rely on, it seems, is "a nice Ear" (299). Unfortunately, the would-be poet who consults a prosodic manual is likely to do so precisely because s/he lacks such a well-developed organ: Johnson's poetaster could have scanned his line on Gildon's system and still have been no closer to understanding that it wasn't a pentameter.

Bysshe's "Rules for making Verses," by contrast, address the practical problem of how to make verses true: how, that is, to vary the prototypical pattern represented by Gray's "The curfew tolls the knell of parting day" while avoiding lines like the one Johnson pillories. Variation from this "norm" (a sequence of ten syllables in which all and only the even-numbered ones are lexically stressed) is more than a matter of aesthetic liveliness: given the realities of stress-distribution in English texts it is a linguistic necessity if the poet's range of expression is not to be savagely crippled. Bysshe's pragmatic focus recalls that of the Elizabethans: the very first account of English metre, George Gascoigne's Certayne Notes of Instruction Concerning the Making of Verse or Ryme in English (1575), was centred like Bysshe's on the problem of acceptable variation. Gascoigne illustrated the difference with the pair 1a and 1b, observing that "in these two verses there seemeth no difference at all, since the one hath the very selfe same woordes that the other hath, . . . yet the latter verse is neyther true nor pleasaunt, and the first verse may pass the musters" (Smith 1:51):

1a. I understand your meaning by your eye. 
1b. Your meaning I understand by your eye. 

A similar interest in the limits of metricality informs the prosodic theorising of his contemporaries.<5> The Elizabethan metrists had little to offer in the way of practical assistance to the versifier, however, because they found the range of permissible variation impossible to codify.<6> It was Samuel Daniel who seems finally to have put a stop to the attempt to theorise the limits of metricality, at least as far as the humanist tradition has been concerned: in rejecting the elaborate and self-contradictory system of regulated foot-substitution proposed by Campion in his Observations in the Arte of English Poesie (1602), Daniel affirmed the mystificatory primacy of intuition over rule-systems, the fetichized "natural" over the supposedly "artificial," which has ever since constituted the only advice traditional metrists can give on the question of making verses true: "euery Versifier that wel obserues his worke, findes in our language, without all these vnnecessary precepts, what numbers best fitte the Nature of her Idiome, and the proper places destined to such accents, as she will not let in, to any other roomes then into those for which they were borne." (Smith 2:378).<7> What we need, it seems, is not rules but "a iudiciall eare," a precept echoed by Gildon and still the best that traditional metrics can offer: "the ear must decide whether the substitution is allowable." (Saintsbury Manual 32).

If the feet and quantities of traditional metrics genuinely and fully represented the structural features of the metre it would be possible to construct a description of metricality with them, for to specify metrical structures is necessarily to define, if only by omission, the nonmetrical. The problem with the traditional model is that what it took from classical Latin was not a theory but the trappings of a theory: it appropriated the notions of syllable-categories, feet and substitution, but it has never succeeded in establishing a set of rules to define and relate them. Thus it cannot tell you how to distinguish the syllable-categories, how objectively to divide a non-standard line into feet, or how to decide whether two feet are equivalent. Hence traditional English scansion is not a precise analytic tool; rather it is a set of vaguely-defined labels for characterising post factum one's intuitions about the rhythm of a line. Where the scansion of a line of French or Latin represents an objective analysis, that of an English line in the traditional system represents nothing but an opinion. Of course, it may permit gesturing between initiates: if one has already acquired a judicial ear -- has already internalized, that is, the tacit system of the metre, as a child internalizes the syntax of its native tongue -- then traditional scansion may allow you to share that opinion with another so endowed, though not in any reliable way. What you cannot do is teach it explicitly as a system in the way you can teach French or Latin or Arabic metrics.

But a code is a code, however its workings may be concealed from casual inspection: in so far as readers and listeners can generally agree on what is or is not clearly metrical even in lines that are new to them (as with Johnson's and Gascoigne's examples), they must be responding to some objective codification of linguistic features in the line itself. It is the failure of traditional metrics to provide an account of this code that has led linguists in the last thirty years to develop a number of so-called "generative" theories of metre. The term comes from mathematics: to generate the members of a mathematical set is to list the conditions governing membership of that set. Thus a generative metric is -- or aspires to be -- an explicit account of the conditions governing metricality. Most generative theories of English metre reject the notion of substitutable feet, and instead see the line as a relation between prosodic form (in terms of underlying, syntactically-determined patterns of phonological stress and juncture) and an invariant abstract metrical pattern or "grid," which in the case of the heroic line consists of ten slots or "syllable positions," labelled alternately Weak and Strong or Odd and Even. Under certain circumstances, roughly corresponding to the traditional environments of elision, syncope, synaloepha and so on, one such position may be filled by more than one actual syllable; under others, an O-position may be empty, producing a catalectic line (though this option is, of course, never exploited in neoclassical versification). Roughly speaking, in generative theories a verse will be true (i.e. metrical) if and only if (a) it can be accommodated to the grid by rules of prosodic adjustment, and (b) no odd-numbered position is occupied by a (variously defined) peak of prosodic prominence. In Halle and Keyser's theory, for example, knife in 2a and -stand in 2b would each be defined as a stress-maximum (i.e. "a fully stressed syllable . . . between two unstressed syllables in the same syntactic constituent within a line of verse" [169]); a stress-maximum in an odd-numbered position will render a line unmetrical. In Kiparsky's theory your knife and -derstand in Even-Odd produce mismatches between prosody and grid (the former would be acceptable in Shakespeare but not in C18 versification); for Magnuson and Ryder, such mismatches are unacceptable tout court:

2a.  Lay your knife and your fork across your plate.
      O   E     O   E    O    E   O  E    O     E
2b. Your meaning I understand by your eye.
     O    E  O   E O  E   O    E  O   E

Although these generative theories are very much in need not merely of development but of radical revision,<8> they nevertheless represent, if not the last word, then the first real advance since the Elizabethans in describing the metrical structure of English verse. What is startling about Bysshe's work is the extent to which it prefigures such theories, a point which I shall elaborate on shortly.<9> It is true that Bysshe talks of "syllables" rather than "syllable-positions," but the mismatch here is only apparent, since elisions for Bysshe (as for his contemporaries) are actual, not notional; if two syllables occupy one position, as in a line like The ling'ring Soul th'unwelcom Doom receives (9), they will be pronounced as one, and the line will remain as decasyllabic in prosodic practice as it is in theory.<10>

Just as Bysshe rejected humanist authority on the question of metrical structure, so also he challenged it in the area of prosodic form. The aim of humanist grammarians like Gildon was not to explore the rich particularity of English, but to discover in it -- or rather, to adapt it to -- the familiar paradigms of academic Latin. In fact this tradition is not yet dead: if you turn to Nesfield, a work still in print and much in demand, you will learn that English nouns have five cases, and that the chief marker of case is "the change of ending that a noun or pronoun incurs" (II.3.10). Not surprisingly, such grammarians treat English word-stress as an analogue of Latin syllabic length. Yet to someone less committed to the universal authority of Latin it must be clear that the analogy is not a good one: stress is a feature of the word (every English polysyllable has just one main stress), whereas length is a feature of the syllable: a Latin polysyllable might have no long syllables at all, or might have three in a row. It is also rather obvious that in English verse two unstressed syllables are not in any sense the equivalent of one stressed syllable. What makes this misidentification even stranger is that Latin has word-stress too, which is clearly (and interestingly) in conflict with syllabic length: the word cano in the first line of the Aeneid, for example, has the first syllable stressed and short and the second unstressed and long.

Of course, Bysshe had to work with the established prosodic categories (nobody yet had challenged these, though the Enlightenment would produce such challenges). He saw, however, that the classical notion of "accent" or pitch-change (a Greek concept irrelevant to Latin) was more applicable to English stress, which (as Fry has shown) is more dependent on pitch-change than on any other cue. Bysshe's three crucial terms, apart from "syllable" itself, are "Accent," "Prevailing Accent" and "Pause." "Accent" Bysshe defines (after Lily) as "an Elevation[, or a Falling (added 1708)] of the Voice on a certain Syllable of a Word" (4).<11> It is clear from this description of Accent as pitch-prominence that for Bysshe it is a relative, not an absolute phenomenon, a feature of the phrase rather than the individual word. Where two stressed syllables are contiguous within the same syntactic constituent (as in none thinks or both lovers in 3 below), one will be subordinated to the other (usually the first to the second) under the stress-rules of English;<12> only the second syllable, therefore, will carry Bysshe's Accent. In fact it seems clear from Bysshe's numerous examples that Accents cannot be adjacent (in this he partly anticipates Halle and Keyser's stress-neutralization rule), so that in 4b the Prevailing Accent in S4 disqualifies both in S5 from carrying an Accent. For the sake of clarity in exposition I will put parentheses around the numbers of positions that contain stresses eliminated as Accents under either branch of what might be called the adjacency rule.

Further testimony to his unconscious understanding of the stress-rules of English can be found in Bysshe's concept of Prevailing Accent, which he defines as "the Accent that is strongest, and prevails most in the first Half-verse";<13>since the "Pause" that terminates the first half-verse (that is, the section preceding the Pause) is necessarily located at the major syntactic juncture in the line, the last tonic syllable preceding it -- by definition the Prevailing Accent -- will tend to be strongest under the Nuclear Stress Rule.<14>

What links Bysshe most strikingly to the generative metrists is the fact that metricality is defined negatively in his system, as in theirs, as the exclusion of prosodic peaks from odd-numbered syllables. For Bysshe, a verse will be true if and only if (a) it can be reduced to ten or eleven syllables by elision, and (b) Accents do not fall upon odd numbered syllables within the line.<15> Thus he observes of the following examples that "tho' the true number of Syllables be observ'd, yet neither of them have so much as the sound of a Verse: Now their disagreeableness proceeds from the undue Seat of the Accent" (6; in the following scansions I underline positions occupied by Accents, bold positions occupied by Prevailing Accents, italicize unmetrically-occupied positions and indicate the pause -- the major syntactic break -- with a colon):

3a. None thinks Rewards render'd worthy their Worth, 
    (1)    2   : 3(4)    5  6   : 7  8   9    10 
3b. And both Lovers, both thy Disciples were. 
    1   (2)   3 4  :  5     6  7  8 9   10

Bysshe's emended versions show that his system will readily entertain what traditionalists call "inversions" or "trochees" (as in positions 5-6 of 4b), provided they follow a major syntactic juncture:

4a. None thinks Rewards are equal to their Worth, 
    (1)    2     3 4   :5   6  7   8   9    10
4b.  And Lovers both, both thy Disciples were. 
     1    2 3    4   :(5)    6  7  8 9   10

In the same way, it will accommodate the so-called pyrrhics and spondees of a traditional scansion, as in 9a. And yet Bysshe's system, while sensitive to the flexibility of neoclassical versification, will weed out the unmetrical lines that have puzzled other writers. Take, for example, the lines cited by Gascoigne above: 5b ("neyther true nor pleasant") has Accent in the seventh position:

5a. I understand your meaning by your eye.
    1 2  3   4  : 5    6  7    8  9   10
5b. Your meaning I understand by your eye
     1    2  3  :4 5  6   7    8  9   10

Or compare Daniel's example 6b, which "wil not be a Verse, though it hath the just sillables," and his metrical version 6a (Smith 2:378):

6a. Though Death doth ruine, Virtue yet preserves
      1     2     3    45  :  6  7   8    9 10
6b. Though Death doth consume, yet Virtue preserves.
      1     2     3    4  5  : (6)  7  8    9 10

Whereas for Daniel it is only "Nature and a iudiciall eare" that will allow you to distinguish them, a Bysshean scansion neatly demystifies the relation between linguistic and metrical form.

To take a more recent example, in a well-known defence of traditional metrics, W. K. Wimsatt and M. C. Beardsley (593) showed how (but failed to show why) a single word-change can render a line of Pope (Essay on Criticism 215) unmetrical.<16> A Bysshean scansion would have solved the conundrum for them:

8a.  A little learning is a dang'rous Thing,
     1  2  3   4   5  :6  7  8    9     10
8b. *A little advice is a dangerous thing
     1  2  3  4  5  :6  7  8    9     10 

Because it takes into account syntax and its effects upon prosodic phonology, Bysshe's system can register distinctions in metrical practice that are invisible to orthodox humanist metrics. The difference between Pope's line 9a (The Rape of the Lock 2.29) and my sabotaged version 9b (unacceptable and unparalleled in neoclassic versification) will not register in a traditional scansion; for a Bysshean, however, 9b is shown to be radically unmetrical by its Prevailing Accent in position 7:

9a. Th'adventrous Baron the bright locks admir'd
       1  2   3    4 5 :  6  (7)    8    9  10 
9b. *The adventurous baron was bright; Locke admired
         1  2    3    4 5   6    7 :   (8)   9  10

By the same token, contrastive accent on was ("he was bright, believe me!") would restore 9b to metricality by returning Prevailing Accent to an even position. To take another example: on a traditional scansion 10b (The Rape of the Lock, 4.35) seems more regular than my emendation 10a (being 60% iambic where 10a is only 40% so), and yet 10b, with its subordinated stress in even position (P4) represents a kind of line that simply doesn't occur in neo-classical versification (though it can be found in Donne and occasionally in Shakespeare):

10a.  On the rich Quilt sinks with becoming Woe
      1    2 (3)    4  :(5)    6    7 8 9    10 
10b. *Sinks on the rich quilt with becoming woe,
       1    2    3 (4)    5  : 6    7 8 9    10

Of course, Bysshe's system as set forth in the "Rules" represents little more than a couple of steps in the right direction: it is still very far from being a full account of English metre. It would be inappropriate to criticise his system specifically as a theory (for its lack of generality, for example, or of explanatory power) since it was only intended as a practical guide to versification; but even on this level there remain many problems. One of these -- understandably, given that English phonology and its relation to syntax have only begun to be systematically investigated in our own century -- is that he does not distinguish the terms Accent and Prevailing Accent as precisely in application as he does in definition, and this confusion might have produced some difficulty for a poet attempting to apply the "Rules". Moreover, the whole question of stress-neutralization and syntactic boundaries, being only implicit in Bysshe's scansions, is insufficiently theorised for a full practical application. But though such a user might never have succeeded in snatching a grace beyond the reach of art, he or she could never have perpetrated the verse cited by Johnson:

11. Lay your knife and your fork across your plate
     1   2     3  :4    5    6  :7  8    9    10 


Whatever its shortcomings, the descriptive system proposed by Bysshe represents (at least as far as heroic verse is concerned) a demonstrable improvement in delicacy, objectivity and practical utility to the versifier over its humanist rival, and yet Gildon's version won the day: something recognizably like it can be found at the back of any standard teaching anthology. It is reasonable to ask why Bysshe's system never succeeded in dislodging or even seriously challenging humanist metrics as the orthodox account, seeming instead to have led a curious subterranean life throughout the eighteenth and nineteenth-centuries in handbooks of "practical versification",<17> only to be independently rediscovered in the 1960s. It is true that as they stand the "Rules" are inimical to the freer versification (and performance practices) of post-neoclassical poetry, but this is not a problem that a thoroughly revised and extended form of the theory could not have addressed. One difficulty is that the theoretical background that would have made such revision and development possible simply did not exist until the advent of generative phonology in the late 1950s. It is symptomatic, for example, that although Newbery's hack sedulously adopts the language of Bysshe's "Rules" he completely fails to understand them, and finally resorts to the appeal to ear: "after all, a person who has a tolerable ear for poetry, will have little occasion for rules concerning the pause or the accents, but will naturally so dispose his words as to create a certain harmony" (12).

A major obstacle to the adoption of Bysshe's system was sheer inertia: the habit of talking about English verse as "Feet compos'd of long and short Syllables" had not only been long established among writers themselves by the beginning of the eighteenth century, but was about to become part of what every schoolboy knew, largely thanks to Gildon, a man who can be credited with initiating the historical process that has turned humanist metrics into part of the educated person's unexamined mental furniture. In English grammars before 1711, the prosodia (where present) consisted of little more than a brief collection of vague and useless generalities: you cannot apply or extend vapidities like "Observatur praesertim numerus syllabarum; non prorsus tamen neglectâ quantitate" (Wallis 127),<18> or "Rhime does consist in a certain Proportion of Feet or Syllables in each Verse or Line, and a Conformity of Sound at the end of several Verses" (Miège 127). Joseph Aickin was, if possible, even less helpful: "The measuring or scanning, of English Verse depends, for the most part, on the equal number of Syllables in sentences: and the agreeing of the last Syllables in a like Sound" (28). There is a perfunctory and conventional air about such formulations, as though the writer were merely observing the forms of grammar-writing by ending with a prosodia. Aickin's grammar, according to R C Alston in his note prefixed to the Scolar Press reprint (1967),

was not, clearly, a popular work since it was printed only once. . . . no English grammar other than that of Wallis (1653) was frequently reprinted before the 'Brightland' grammar of 1711, and it was not until the second quarter of the eighteenth century that the teaching of English grammar became an essential part of the grammar school curriculum.

Gildon and Brightland's Grammar of the English Tongue . . . to which is added A New Prosodia . . . All adapted to the Use of . . . the Schools of Great Britain, was certainly popular, being re-issued nine times between 1711 and 1782, and it is in this primer that we find for the first time a teachable humanist metric of English (by "teachable" I mean that it is sufficiently defined and structured to work as a game, and thus to constitute a source of classroom transactions). The authors give rules of a kind for "quantity," which they identify in a relatively straightforward way with what traditional metrists now call stress, and go on to provide a recognisable sketch of the now-familiar schemes of foot-substitution. But if Gildon's grammar sowed the first seeds of traditional classroom metrics, it was Lindley Murray that began the intensive cultivation. Murray's famous Grammar (1792), which "for a long time was used in schools to the exclusion of all other grammar books" (DNB), was immensely influential; it "went through at least fifty editions, it sold millions of copies in both England and America, and was freely imitated by scores of later grammar writers" (McLaughlin 84). Murray presents, in his methodical fashion, the entire paraphernalia of practical traditional metrics, with handy illustrative scansions.

Thus what we now think of as traditional metrical terminology entered the culture by default, as it were, on the coat-tails of traditional grammar, another scheme based not on an analysis of English structures but on an appropriation of Latin terms. For the last two centuries, humanist grammar and humanist metrics have formed a pair of mutually supporting paradigms of commonsense knowledge -- Foucauldian "discourses" standing squarely in the way of change. Though the scale is vastly different, the comparison with Kuhn's favourite paradigm-shift from Ptolemy to Copernicus/Kepler is instructive: before Copernican ideas could become anything other than manifest nonsense to non-mathematicians, a whole battery of fundamental transformations had to take place in commonsense perceptions of many matters, from the physics of moving objects to the place of humanity in the universe. In the same way, widespread acceptance of a rational linguistically-based system of scansion for English would require a major overhaul in people's everyday assumptions about language, metre, the relation between speech and writing, and so on. It might be supposed that the practical superiority of linguistic metrics would prejudice people in its favour, but practical utility seems curiously unimportant in such matters. Most readers of English verse probably care little about the inadequacies of traditional metrics as a tool of aesthetic description (let alone of taxonomy or stylometry), and never discover its complete failure as a grammar of the metre, because if they learn to recognise and produce metrical form they do so not by reference to a manual but by unconsciously internalising the grammar of the metre.

Traditional metrics "works," therefore, in the crude sense that it doesn't actually get in the way of an intuitive understanding of the metre, and may even permit some communication between initiates, rather like the impenetrable metaphors of wine-buffs. But like such metaphors, and unlike the explicit metrical accounts of Latin or French, it cannot explain anything to those who don't already in some sense understand. It is, in other words, a language for insiders. As William Hand Browne wrote in 1889, "It certainly is an odd thing that while everybody enjoys good verse, and nearly everybody can make very tolerable verses, at least so far as metre and rhythm are concerned . . . hardly anywhere can we find a lucid and intelligent explanation of the principles on which English verse is constructed" (97). We may be certain that "everybody" here does not include the man on the Clapham omnibus. In literate societies metre, like other cultural systems (including language itself), tends to reflect social hierarchies. One often finds a dichotomy between a spontaneously acquired demotic system of versification with fairly simple and overt linguistic regulation on the one hand, and on the other a more subtly regulated and less immediate system acquired only by the educated. In Latin, for example, there is a broad gulf between the loose Saturnian stress-metre of domestic charms and lullabies and the quantitative hexameters of Ovid and Vergil, metrically unintelligible to the uneducated native speaker (Allen 126-7n.); the precisely-defined organization of the classical alexandrine is similarly far removed from that of the verses in a French Christmas card. Similarly, whereas in English the popular four-beat metre of nursery rhymes, protest chants, football incantations, advertising jingles and so on is available to every native speaker, whatever their class or educational attainment, the capacity to recognise, perform and produce so-called accentual-syllabic verse has -- at least since the demise of popular Elizabethan theatre -- been confined to the educated, and has thus in the past had something of the function of a social shibboleth, another way of sorting out the right sort of people who know what to do with a fish-fork and which way to pass the port. Thus Saintsbury, for example, appeals not to any theory to justify his scansions but rather to the practice of "[e]very well-educated and well-bred Englishman, who has been accustomed to read poetry and utter speech carefully" (Manual 21). Having an "ear" for metre is thus rather like having "taste" or "breeding": such language fetichizes a particular internalized code into a seemingly natural distinction of social class. It follows that the failure of traditional metrics to help you to tell good lines from bad is precisely what makes it preferable to any genuine attempt to account for the metre in linguistic terms: a shibboleth, after all, is no longer a shibboleth once blighters like Bysshe make the answers available to any old Tom, Dick or Harry. If Johnson's versifying bore had known his Bysshe, for example, we would all have been deprived of the pleasure of looking down our noses at him.

But why hang on to the terms at all, if they are virtually without practical utility? No doubt sheer cultural inertia is a factor here, but inertia need not be the whole explanation for the perpetuation of anachronistic cultural practices: seemingly obsolete traditions may belie their practical inutility and irrelevance by fulfilling an important symbolic function. Just as traditional metrics is only the trappings of a theory, for example, the British monarchy now represents only the trappings of a system of government, irrelevant to the practical business of legislation and administration and clearly inappropriate as a method of selecting the head of state in a democracy. Yet the monarchy has (at least until its recent farcical Entzauberung) performed an important ideological function in mystifying and dignifying the naked power of the state. In the same way, without the learned paraphernalia of traditional metrical description heroic verse would appear worryingly similar to its demotic counterpart: spontaneous, "artless", apparently uncoded. So the role of traditional metrical terminology in the institution of English literature may be essentially a ceremonial one, drawing upon the ancient pedigree of the terms; it may function as a badge of status, that is, signifying the elevated literary nature of heroic metre and insisting on its difference from the apparently natural or spontaneous forms of popular verse.<19> It is, perhaps, no coincidence that textbooks of humanist metrics seem to have flourished most in the half-century following the 1870 Education Act, which produced a new kind of secondary school, one largely indifferent to the classical traditions of Harrow and Winchester. When Joseph Mayor, for example, offers to "those who have not had a training in metre through the practice of Greek and Latin versification" "a useful substitute for the latter in schools of the more modern type" (vi), his intention is to smuggle in a humanist curriculum (or part of it) through the back door. Though he could scarcely have been aware of it, his aim was at bottom a political one: to civilise and assimilate the threatening mass of new readers, channelling their literary appetites away from divisive social speculation and towards the canons of high bourgeois culture -- or, at any rate, "a useful substitute" for it.


1. It ran through eight editions or reprints in its first forty years. Admittedly its popularity -- even Blake possessed a copy -- may have owed something to the accompanying dictionary of rhymes, and perhaps even more to Bysshe's collection of passages for imitation from the English poets (see Dixon, Connaughton).

2. Alexander Pope, Epistle to Dr Arbuthnot (1735), line 151. See also the Dunciad (1728), 3.167-72.

3. This requirement is categorical for the neoclassical versification of Bysshe's own period, and it is normative for so-called accentual-syllabic metre in all periods. See, for example, Dorothy L. Sipe, Shakespeare's Metrics (Yale Studies in English 166; New Haven, Connecticut: Yale University Press, 1968).

4. "The number of syllables is thus the sole determinant of the structure of our verses."

5. See, for example, Puttenham, Arte of English Poesie (1589), Smith 1:1-193; James VI of Scotland, Ane Schort Treatise Conteining Some Reulis and Cautelis to be Observit and Eschewit in Scottis Poesie (1584), Smith 1:208-226; Thomas Campion, Observations in the Arte of English Poesie, Smith 2:327-355; Daniel, A Defence of Rhyme, Smith 2:356-384.

6. The immediate reason for this was the wildly unstable signification of the crucial term "length" in their writings: a syllable might be designated "long" by virtue of lexical stress, orthography, contrastive accent, vowel-length, etymology or even personal fiat (see Attridge, Well-weighed Syllables, for a detailed account of this).

7. For further discussion, see Hiller and Groves.

8. For discussion of the approach see Bowley, Attridge, and Groves; for a radical revision, see Groves.

9. None of the generative metrists reveals any knowledge of Bysshe's work; it is interesting, however, that the Port-Royal grammarians, with their interest in universal grammar, were an important influence in Chomsky's intellectual development.

10. It is a sign of Bysshe's serious intellectual purpose, incidentally, that his section on elision more than doubled in size in the 1718 edition. The mere hack would not have bothered to expand it (it would be hardly likely to improve sales vis-à-vis Gildon).

11. Compare Bysshe's simple and sensible formulation with a contemporary's baffling attempt to reproduce in English the full panoply of classical distinctions (with typical confounding of orthography and pronunciation): "There are three Accents, 1 the heavy or short accent descending towards the right thus [`] is the sign of a Syllable elevated above the rest, as còntrary. 2. The sharp or long Accent, riseing to the right, thus [´] is the sign of a Syllable made long, mét; when both these Accents meet in one Syllable, it is called (3) a Circumflex as contrîve" (Aickin 25-6).

12. Where two words belong to the same phrase the first stressed syllable is subordinated in stress to the second under what Chomsky and Halle call the Nuclear Stress Rule (89-91).

13. Again an echo of Lancelot, this time from his "Breve instruction de la poësie Italienne" (79): "Cette inegalité‚ dépend de l'accent qui domine dans l'hemistiche. Car la cesure est toûjours … la fin du mot où se trouve cet accent." The issue, of course, is not one of literary etiquette but of the relevance of what Bysshe wrote to English phonology and metrics.

14. The Pause "must be observ'd at the end of the Word where such [i.e. Prevailing] Accent happens to be, or at the end of the following Word" where that is an enclitic that is, "a Monosyllable that may be sunk in its Pronunciation, and whose Construction is Govern'd by that word on which the Accent is; as, Despise it And more noble Thoughts pursue." (4). In the 17 examples given to illustrate Pause, it always occurs at the major intonational phrase boundary in the critical area (i.e. between the third and ninth syllables); thus in the following line, Bysshe rejects the idea of placing the Pause after the fourth syllable rather than the seventh, since "tho' it be no Violence to the Ear, yet it is to the Sense" (6):

          Mirrors are taught to flatter, but our Springs
           1  2   3    4      5   6  7    8  9      10 

15. Specifically "the 3rd, 5th or 7th" (7); he specifies only these three because his rules restrict the Pause to a zone extending from the second to the eighth syllables (4), which effectively eliminates the first and ninth positions from the ban on Accent (thus permitting initial reversals).

16. "all we have changed is the position of one relative accent [within the foot], which makes it impossible that the syllable 'is' should receive a stronger accent than the preceding syllable, and hence impossible that there should be five iambs in the line." Requiring five iambs in every line is, it will be recalled, one of the absurdities laid to the charge of Bysshe.

17. Take, for example, Tom Hood [the Younger]'s The Rules of Rhyme (1869), which prints in an appendix a treatise on versification it claims to be founded on something called "The Young Poet's Guide." Hood writes at some length in orthodox Bysshean terms of Accent and Pause, noting that "the seat of the Accent" is syntactically determined, depending "mainly on the sense; and on the sense, moreover, of each individual line." (163). The Rules of Rhyme perpetuates a subversive counter-tradition and demonstrates that something like an alternative anti-academic Bysshean metrical discourse was still circulating as late as the 1860s; in fact Bysshe's Rules were reprinted verbatim as an appendix to a later edition of Hood's work, the Practical Guide (1877).

18."The chief thing to be observed is the number of syllables (without, however, completely disregarding quantity [syllabic length])."

19. A similar dichotomy obtained in Renaissance humanist thought between the prestigious coded and artful metrum of classical Latin verse and the merely spontaneous "natural" rhythmus of vernacular poetry including, ironically, heroic verse. See Hiller and Groves for a fuller discussion of this.


Aickin, Joseph. The English Grammar. London, 1693.

Allen, W. S. Vox Latina: A Guide to the Pronunciation of Classical Latin. 2nd ed. Cambridge: Cambridge University Press, 1978.

Attridge, Derek. Well-weighed Syllables: Elizabethan Experiments in Quantitative Verse. Cambridge: Cambridge University Press, 1974.

--. The Rhythms of English Poetry. London: Longmans, 1982.

Boswell, James. Life of Samuel Johnson. Ed. George Birkbeck Hill, revised and enlarged edition by L. F. Powell. London: Clarendon Press, 1934.

Bowley, C. C. "Metrics and the Generative Approach." Linguistics 121 (1974): 440-448.

Brogan, T. V. F. English Versification 1570-1980: A Reference Guide with a Global Appendix. Baltimore: Johns Hopkins UP, 1981.

Bysshe, Edw[ard], Gent. The Art of English Poetry: Containing, I. Rules for making Verses. II. A Dictionary of Rhymes. III. A Collection of the most Natural, Agreeable and Noble Thoughts, viz. Allusions, Similes, Descriptions and Characters, of Persons and Things; that are to be found in the best English Poets. London, 1702.

Chomsky, Noam, and Morrris Halle. The Sound Pattern of English. New York: Harper & Row, 1968.

Connaughton, Michael. "Richardson's Familiar Quotations: Clarissa and Bysshe's Art of English Poetry." Philological Quarterly 10 (1981): 183-198.

Culler, A. Dwight. "Edward Bysshe and the Poet's Handbook." PMLA 63 (1948): 858-885.

Dixon, Peter. "Edward Bysshe and Pope's "Shakespear ," Notes and Queries 11, n.s. (1964): 292-293.

Fry, D. B. The Dependence of Stress Judgments on Vowel Formant Structure." Proceedings of the Fifth International Congress of Phonetic Sciences. Münster: n.p., 1965. 306-11. Rpt. in Acoustic Phonetics, ed. D. B. Fry. Cambridge: Cambridge University Press, 1976, 425-30.

Fussell, Paul. Theory of Prosody in Eighteenth-century England. Connecticut College Monographs 5. New London: Connecticut College, 1954.

[Gildon, Charles]. A Grammar of the English Tongue, with Notes Giving the Ground and Reason of Grammar in General. To which is added, A New Prosodia; or, The Art of English Numbers. All adapted to the Use of Gentlemen and Ladies, as well as of the Schools of Great Britain. London: Printed for John Brightland, and sold by Mr. Guy. 1711.

Gildon, Charles. The Complete Art of Poetry. London: 1718.

Groves, Peter L. Strange Music: the Metre of the English Heroic Line. ELS Monograph Ser. 74. Victoria, B.C.: English Literary Studies, 1998.

Halle, Morris, and Samuel J. Keyser. "Illustration and Defence of a Theory of the Iambic Pentameter." College English 33, 1971:154-76.

Hiller, Geoffrey G., and Peter L. Groves. Samuel Daniel: Selected Poems and "A Defence of Rhyme." Asheville, NC: Pegasus Press, 1998.

Hood, Thomas [the Younger]. The Rules of Rhyme: A Guide to English Versification, With A Compendious Dictionary of Rhymes, An Examination of Classical Measures, and Comments Upon Burlesque, Comic Verse, and Song-writing. London [1869].

--. Practical Guide to English Versification. To which are Added, Bysshe's 'Rules for making English Verse'. New and enlarged edn. London, 1877.

Kiparsky, Paul. The Rhythmic Structure of English Verse." Linguistic Inquiry 8 (1977): 189-247.

Lancelot, Claude. Quatre traitéz de poësies, latine, françoise, italienne, et espagnole. Paris: 1663.

Magnuson, Karl, and Frank G. Ryder. Second Thoughts on English Prosody." College English 33 (1971): 198-216.

Mayor, Joseph B. A Handbook of Modern English Metre. 2nd ed. Cambridge: Cambridge UP, 1912.

McLaughlin, John C. Aspects of the History of English. New York: Holt, Rhinehart and Winston, 1970.

Miège, Guy. The English Grammar. London: 1688.

Murray, Lindley. The English Grammar, Adapted to the Different Classes of Learners. With an Appendix, Containing Rules and Observations for Promoting Perspicuity in Speaking and Writing. York: 1795.

Newbery, John. The Art of Poetry on a New Plan. London: 1762.

Nesfield, John Collinson. Outline of English Grammar for Home and Office. Rev. edn. London: Macmillan, 1982.

Omond, T. S. English Metrists: Being a Sketch of English Prosodical Criticism from Elizabethan Times to the Present Day. Oxford: Clarendon Press, 1921.

Saintsbury, George. A History of English Prosody, from the Twelfth Century to the Present Day. 2nd ed. London: Macmillan, 1923.

--. A Historical Manual of English Prosody London: Macmillan, 1930.

Sampson, George. The Concise Cambridge History of English Literature. Cambridge: Cambridge UP, 1970.

Smith, G. Gregory, ed. Elizabethan Critical Essays. London: Oxford UP, 1904.

Wimsatt W. K., and M. C. Beardsley, "The Concept of Meter: An Exercise in Abstraction," PMLA 74 (1959): 585-598.

ISSN 1546-0401