What evidence is there for the "recession of accent" theory?

Question

In the late 19th and early 20th century, there was a theory that certain oddities in the rhythms of Shakespeare and other early modern English poets could best be explained by recession of accent.

The history of this theory seems to be as follows. In the 1870s, Alexander Schmidt compiled a collection of examples of “Changeable accent of dissyllable adjectives and participles” in Shakespare’s iambic verse. By this he meant two-syllable words which Shakespeare placed starting on an odd syllable in one line, and on an even syllable in another, so that in order to scan both lines as perfect iambic pentameter, the word has to be pronounced with different stress in the two instances. Schmidt’s examples include:

Therefore my verse to constancy confined, [Sonnet 105]
Supposed as forfeit to a confined doom. [Sonnet 107]

He means, my lord, that we are too remiss. [Richard II III.2]
That thus we die, while remiss traitors sleep. [I Henry VI IV.3]

Whilst thou liest warm at home, secure and safe. [Taming of the Shrew V.2]
Upon my secure hour thy uncle stole. [Hamlet I.5]

_{Alexander Schmidt (1875). Shakespare-Lexicon, volume II, p. 1413–1415. London: Williams and Norgate.}

Schmidt concluded:

And thus it may be stated as a general rule, that dissyllabic oxytonical† adjectives and participles become paroxytonical‡ before nouns accented on the first syllable.

_{Schmidt, p. 1413.}

_{† stressed on the last syllable ‡ stressed on the penultimate syllable}

In his influential Milton’s Prosody (1893), Robert Bridges described this as the “rule of the recession, or retreating of accent” (p. 13), and in a detailed discussion of the rule, he argued for an expanded definition:

The only fault to find with [Schmidt’s] definition is that it confines recession of accent to adjectives and participles, and that it requires the determining (following) word to be a noun, which it is true that it is in all his instances; but I should be inclined to consider this unessential, and treat the recession as being due entirely to collision of accents. I do not see that it has anything to do with the sense, or with the adjective being in the predicate or not, as implied by Dr. Schmidt’s remarks.

_{Robert Bridges (1893). Milton's Prosody, p. 54. Oxford University Press.}

The theory continued to be accepted at least into the 1920s:

The deliberate violation of normal word-accent to fit the metrical stress […] is not such an entirely arbitrary procedure as it might at first seem; for at one period in the history of the language the accent of many words (especially those of French origin) was uncertain. Chaucer could say, without forcing, either náture or natúre. The revival of English poetry in the sixteenth century owed a great deal to Chaucerian example, and thus a tradition of variable accent was accepted and became practically a convention, not limited to those words in which it had originally occurred. […] The wrenching of accent for metrical purposes, moreover, is not confined to the dissyllabic words which show the simple recession of accent.

_{Paull Franklin Baum (1922). The Principles Of English Versification, p. 190. Harvard University Press.}

However, what I have been unable to find, is the body of evidence for the theory, beyond the pairs of lines from Shakespeare collected by Schmidt, and other examples of lines that are difficult to scan as iambic verse. The theory seems to rely on the assumption that the scansion is given, so that we can work out the pronunciation of the words from the scansion. But absent an independent body of evidence, we have no justification for proceding this way: it makes just as much sense to take the pronunciation of the words as given, and work out the scansion from the pronunciation. So the theory, at least insofar as it is presented by the authors I have quoted here, seems unconvincing.

I find in Richard Bentley’s 1732 edition of Milton’s Paradise Lost, the following note on Milton’s line “Encamp their Legions; or with obscure wing”:

I suspect, He gave it, with more Art, and smoother Accent, “Encamp their Legions; oft with wing obscure”

_{Richard Bentley (1732). Milton’s Paradise Lost, note to book II, line 132. London: Jacob Tonson.}

Under the “recession of accent” theory the accent was smooth in the original line so that Bentley had nothing to object to, unless the practice had already died out the early 18th century.

So my questions are,

1. What is the modern consensus, if any, on the “recession of accent” theory?
2. Is there a body of independent evidence for the theory?
3. If so, what is it? If not, how was the record corrected?

Answer 1

TL;DR: The question remains open, with little progress having been made in the last century. Modern commentators on the issue mostly take a neutral or skeptical position, or suggest that the phenomenon might apply in some cases but not others. Evidence from Richard Bentley, however, shows that the phenomenon, if it ever existed, had disappeared by the late 17th century.

Modern commentators

Whether recession of accent was in fact intended in cases like those before us is an issue that, a hundred years later, has still not been settled. Kiparsky† has argued pretty convincingly in favour of recession in cases of this kind; Attridge,‡ after discussing the matter at some length, remains largely non-committal. I am inclined to feel that recession was intended, at least in the two cases from Milton.§

_{David Keppel-Jones (2001). Strict Metrical Tradition: Variations in the Literary Iambic Pentameter, p. 84. McGill-Queen’s University Press.}

_{† Paul Kiparsky (1975). ‘Stress, Syntax, and Meter.’ Language 51, pp. 575–616. Kiparsky uses the term “Rhythm Rule” rather than “recession of accent”.}

_{‡ Derek Attridge (1982). The Rhythms of English Poetry. London: Longman.}

_{§ I am not sure that we can trust Keppel-Jones’ inclination. The “two cases from Milton” that he cites are I.406 and I.735, both of which were objected to by Richard Bentley, as discussed below.}

Sprott† denies the existence of stress-final pairing (in his terms, adjacent stresses within a single foot), and is able to allow adjacent stresses only if they fall in separate feet. Consequently, he is driven to suggest implausibly that XII.409 should be spoken: ‘ImPUTed BEcomes THEIRS by FAITH, his MERits’ with recession of stress on ‘becomes’ when there is clearly a stress-final pairing: ‘Imput-ed be-COMES THEIRS’. This is an extreme instance of the tendency to evade paired stresses by positing ‘recession of accent’, a phenomenon never certainly present in Milton and much rarer earlier than some prosodists have envisaged.

_{John Creaser (2007) ‘“Service Is Perfect Freedom”: Paradox and Prosodic Style in Paradise Lost’. The Review of English Studies, New Series 58:235, p. 277.}

_{† Samuel Ernest Sprott (1953). Milton’s Art of Prosody. Oxford University Press.}

(14) a. A maid of grace and complete majesty (Shakespeare, Love’s Labour’s Lost I.1)
b. His means of death, his obscure funeral (Shakespeare, Hamlet IV.5)
c. Sir, in good sooth, in sincere verity. (Shakespeare, King Lear II.2)
d. The divine property of her first being (Milton, Comus 469)
e. In profuse strains of unpremeditated art. (Shelley, ‘To a Skylark’ 6)

The examples in (14), which have been copied from Bridges 1921,† are frequently cited in discussions of English meter. Bridges attributed these to “recession of accent…not now heard” (p. 67). As the main (sole?) evidence for the accent recession is the scansion of the lines, Bridges’s proposal is not compelling, especially since accent movement has not been adduced by Bridges or anybody else as an explanation for the similar line-initial examples […]. In view of this there is reason to question accent recession as an explanation for the facts in (14).

_{Morris Halle (2008). ‘On Stress and Meter and on English Iambics in Particular’. In Kristin Hanson and Sharon Inkelas, eds. (2008). The Nature of the Word: Studies in Honor of Paul Kiparsky. MIT Press.}

_{† These are examples (4), (38), (47), (66), and (110) respectively, in Robert Bridges (1893). Milton’s Prosody, pp. 55–62. Oxford University Press.}

The double iamb

The competing theory to “recession of accent” is that the “double iamb” (two unstressed syllables followed by two stressed syllables) was considered an admissible variation in iambic verse by many English poets. The double iamb theory has three features that make it more attractive than the “recession of accent” theory: it does not require us to hypothesize a feature of English speech that happened to disappear just before the first pronouncing dictionaries appeared; it does not depend on the number of syllables in the words, and so explains a wider variety of examples; and it applies to late modern verse (where we know recession of accent does not work) just as well as it does to early modern verse.

The double iamb does not seem to have been properly recognized until the mid-20th century. At the beginning of the century, the phenomenon was doubtful enough for C. M. Lewis to prefix it with a cautionary disclaimer, “Metrists usually say”:

Metrists usually say that the pyrrhic is sometimes admitted in iambic verse; and some leading authorities point out that it generally is found with a spondee immediately afterwards.

_{C. M. Lewis (1907). The Principles of English Verse, p. 42. New York: Henry Holt.}

Robert Bridges noted it as “a common irregularity” in Milton, without seeming to realize that it covered all the cases that he put under the heading of “recession of accent” too.

The matter stands thus: in all Milton’s verse there is a frequent occurrence of the following rhythm, that is, a foot of two unstressed short syllables preceding a foot composed of two heavy syllables, as in these lines from Shakespeare’s Midsummer-Night’s Dream,

The ploughman lost his sweat, and the green corn.
Before milkwhite, now purple with love’s wound.
Love takes the meaning in love’s conference.

It is common in Milton’s early verse, which is much influenced by the verse of Shakespeare’s first style; and he always made use of it. Whatever the account may be, it is pleasant to the ear in the smoothest verse, and is so, no doubt, by a kind of compensation in it. In typical cases there is no possibility of stress in the first short foot, and the first heavy syllable of the next foot seems to carry what has been omitted, with an accentuation bearing relation to the sense.

Instances occur everywhere in Milton.

_{Robert Bridges (1893). Milton's Prosody, p. 58. Oxford University Press.}

The double iamb was, I think, first stated in the form of a rule by American poet John Crowe Ransom:

These are substantially the exceptions as codified by Bridges in The Prosody of Milton [sic], the best handbook we have on iambic pentameters. What Bridges codified was Milton’s code, as it had been for several generations the code of Milton’s predecessors, and would be for his successors over a century and a half; since then it has been well known to poet-prosodists, and adhered to systematically when they pleased. But it is not quite complete, in my judgment. I wish Bridges had added:

4. Any two successive iambic feet might be replaced by a double or ionic foot.

_{John Crowe Ransom (1956). ‘The Strange Music of English Verse’. The Kenyon Review 18:3, pp. 470-471.}

The name “double iamb” was not introduced until the 1990s:

We should drop the pyrrhic foot and accept in its place the double-iamb, as one of the six foot-terms necessary: iamb, trochee, anapest, dactyl, spondee, double-iamb.

This change was suggested by John Crowe Ransom […] Ransom uses the Greek term “ionic foot”, but plain English seems preferable, carrying with it a reminder that a double-iamb counts as two feet, and having the advantage of suggesting that this pattern is quite normal to the iambic base.

_{Robert Wallace (1993). ‘Meter in English’. In David Baker, ed. (1996). Meter in English: A Critical Engagement. University of Arkansas Press.}

Bentley’s edition of Milton

The difficulty in evaluating the “recession of accent” theory versus the “double iamb” theory is finding somewhere solid to start: we can’t simultaneously deduce the pronunciation from the scansion and the scansion from the pronunciation. But one possible place to start, as suggested in the question, is Richard Bentley’s 1732 edition of Paradise Lost. This is usually derided for the editor’s attempts to correct what he sees as the mistakes of Milton (see my discussion here). But because Bentley did not realise that correcting Milton’s verse was beyond his poetic powers, we can make use of his edition to identify cases where Milton’s rhythm did not seem “smooth” to the editor and so needed “correction”.

I went systematically through the first two books of Paradise Lost collecting lines containing double iambs, and where “recession of accent” would make the line scan as perfect iambic pentameter, and then I checked Bentley’s edition to see what, if anything, he says about them. I found seven such cases, and in all but one of them Bentley objects to the line and suggests an alternative that rearranges words to avoid the double iamb.

Line	Milton’s text	Bentley’s emendation
I.238	Of unblest feet. Him followed his next Mate,	Of feet unbless’d. Him follow’d his next Mate,
I.406	Next Chemos, th’obscene dread of Moabs Sons,	Next Chemos, Dread obscene of Moab’s sons,
I.735	And sat as Princes, whom the supreme King	And sat as Princes: whom the King supreme
II.7	Thus high uplifted beyond hope, aspires	—
II.132	Encamp thir Legions, or with obscure wing	Encamp their legions; or with wing obscure
II.210	Our Supream Foe in time may much remit	Our Foe Supreme in time may much remit
II.615	In confus’d march forlorn, th’ adventrous Bands	In march confus’d forlorn th’adventurous Bands

This, I think, is evidence that “recession of accent”, if it had ever existed, had died out by the time Bentley (born 1662) was learning English. But that was the same period of time in which Milton was writing Paradise Lost, and so it seems unlikely that Milton relied on the phenomenon either. But then why did Bentley object to Milton’s scansion in these lines? The straightforward answer would seem to be that Milton considered the double iamb to be an admissible variation in English iambic verse, whereas Bentley (a classicist) did not. Additional support for this theory comes from the fact that in a few cases Bentley offers “corrections” to Milton’s double iambs in lines where “recession of accent” would not apply because the paired stresses are on single-syllable words:

Line	Milton’s text	Bentley’s emendation
II.261	Thrive under evil, and work ease out of pain	Thrive under evil, and out of pain work ease
II.818	And my fair Son here showst me, the dear pledge	And My fair Son here shew’st me; dearest pledge
II.955	Undaunted to meet there what ever power	Undaunted, there to meet whatever Pow’r

(In many more cases Bentley does not offer corrections to Milton’s double iambs, so it is hard to be sure exactly what Bentley’s theory of English prosody was, if he had one. However, his habit of criticizing the rhythm of a line only in tandem with offering a correction means that he may simply have remained silent in cases where he did not have a good enough suggestion. For example, of the seven lines I gave previously, II.7 “Thus high uplifted beyond hope” is the only case in which a simple rearrangement of words fails to regularize the line. In this line a rephrasing is needed, for example, “Thus high upraised beyond all hope”, but this would repeat “raised” from II.5, something which Bentley also frowned on, and “all” is an expletive. So it is possible that he simply gave up on attempting to correct this and other lines.)