LINGUIST List 14.2625

Tue Sep 30 2003

Review: Historical Ling/Phonology: Minkova (2002)

Editor for this issue: Naomi Ogasawara <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in.

If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Simin Karimi at


  • Stephen Laker, Alliteration and Sound Change in Early English

    Message 1: Alliteration and Sound Change in Early English

    Date: Mon, 29 Sep 2003 18:14:37 +0000
    From: Stephen Laker <>
    Subject: Alliteration and Sound Change in Early English

    Minkova, Donka (2002) Alliteration and Sound Change in Early English, Cambridge University Press, Cambridge Studies in Language 101.

    Announced at

    Stephen Laker, Leiden University


    Despite an ever-increasing production of books and articles on the metre of especially Old English alliterative verse, the study of the alliterations themselves, their changing developments both within Old English and, more prominently, into Middle English, has received little attention. Largely, we are still reliant on Schumacher's Bonn dissertation of 1914: a concise, but by modern standards, linguistically unanalysed compendium of the many alliterative possibilities found in the Middle English verse corpus. That after so long a period of disregard the study of English historical phonology as reflected in the changing nature of the verse alliterations is being addressed, can only be welcomed.


    Chapter 1 (pp. 1-21) begins with a useful introduction to the Germanic background of the Old English alliterative tradition and its revival in Middle English in the West (the question why it was revived only in the West is not touched upon). Chapter 2 (pp. 22-70) then goes on to outline Minkova's ideas as to the versification of Old English. The approach follows that of Stockwell and Minkova (1997) which is (p. 35): ''a new synthesis of Sievers' theory that incorporates insights found in Bliss (1967) and Cable (1974), but it follows neither of them in all details''. An interesting aspect of this theory is the notion that resolution had become a poetic artifice very early on (e.g. as compared to High German, cf. Vennemann 1995) and was not understood by the linguistic intuitions of native Old English speakers (p. 7, 19, 40). This is also interesting because throughout the book the reader is constantly reminded (p. xv, 6, 7, 18 etc.) of how alliterative compositions are otherwise a valuable resource for the reconstruction of the phonological properties of contemporary languages. The chapter concludes with a brief outline of Middle English stress patterns and versification.

    The analysis of alliterations begins in chapter 3 (pp. 71- 134). Generations of students of Old English have learnt of the special poetic licence given to alliterations on the velars k- and g-. The general view on the voiceless velar k- is that though it was fronted and subsequently affricated before front vowels (not mentioning other positions), it could still alliterate with non-affricated k-. This essentially means that in verse Old English _cynu_ 'kin' alliterates with _cinu_ 'chin', despite their assumed onset variation as in Modern English. This standard handbook view is rejected by Minkova who does not believe such an ''eye-alliteration'' was at play. Instead it is argued that k- was indeed palatalized to kj-, but not affricated, thus retaining its velar properties in alliteration. This resembles the reconstruction already assumed for Early Old English. Due to i-umlaut, a further palatal environment in English arose, however, which remained distinct from that of the earlier palatalization environment, and only caused a phonetic variation, e.g., as in the palatal and non-palatal variations of /k/ in the (Old) English words _cyning_ 'king' [k'] and _corn_ 'corn' [k]. Phonetically this gives for Old English three distinct reflexes of Germanic *k: [k], [kj], [k'], cf. (p. 106). An interpretation, Minkova adds (p. 99), which is theoretically possible, drawing on research carried out by Keating and Lahiri (1993).

    But the situation is complicated somewhat by the assumption of yet another reflex of Germanic *k in Old English: the presence of the affricated pronunciation [ts^] (s^ stands for the voiced palato-alveolar fricative as in English <sh->), namely word internally and in final absolute position (p. 112) as in Old English _dic_ 'ditch'. Minkova defends this proposal of positionally- governed affrication (p. 110): ''An argument can be made that the palatalized velars became affricated first word- internally, where they could appear in coda position, a prototypical position of neutralization.'' Though not explicitly spelled out, the main need for the assumption of positionally-governed affrication is to accommodate Old English <-c(c)-> spellings, i.e. as in <fecc(e)an> 'to fetch' < *fetjan, where an etymological dental stop undergoes affrication before a following /j/, cf. p. 110. Hogg (1992) assumes such spellings, which are abundant in Late West Saxon, to be in use by at the latest the beginning of the ninth century. But we must logically assume that in order for the spelling convention to be adopted <c> must already have been able to express something equivalent to [ts^] in the Old English orthographical system prior to its transferred use in the -tj-forms. This is too early. The only way therefore of accommodating the <-c(c)->-forms while assuming that Old English verse was still alliterating on velars is then to posit affrication according to position. One would have liked, however, to have seen some typological evidence for this thesis with regard to palatalization and affrication processes in other languages; more especially, similar contemporaneous palatalizations in other Germanic languages. A closely related North Sea Germanic language - Old Frisian - does in fact display such a positionally- governed palatalization and affrication, but regrettably for Minkova's thesis in precisely those positions she predicts are less susceptible to it, namely before high vowels but not after them, i.e. as a anticipatory assimilation process, cf. Nielsen (1985) - Nielsen also gives some place-name evidence for the same positionally- governed palatalization in dialects of Low German. These complications in Old Frisian and Low German ought really to have been addressed.

    Next the problem of voiced velar g-alliterations is tackled (pp. 113-121). In contrast to later verse it has been observed that Old English reflexes of the Germanic voiced velar fricative *g^ (g^ represents the IPA gamma sign) as in Old English _gold_ 'gold' alliterate with reflexes of the Germanic approximant *j (e.g. _geong_ 'young') despite a breaching of phonemic boundaries. Minkova makes a plausible argument that the crossing of phonemic boundaries in early verse may be explained when the voiced velar fricative articulation /g^/ is assumed for Early Old English. This would have had both palatal and non-palatal allophones. A valid argument is made that the palatal-fricative allophone [g^'] (curly-tail j in IPA) would have been so phonetically similar to the palatal approximant /j/ for such a breach of phonemic boundaries to be tolerated. However, it would also have represented an allophonic bridge through which the non- palatal voiced velar fricative [g^] could alliterate too. That such alliterations are no longer observed in later verse would indicate a change in this practice, however. The change would result from the occlusion of the non- palatal voiced velar fricative [g^] to [g]. By this interpretation, dating of the manuscript evidence could provide an approximate terminus post quem (c. 950) for the occlusion in Old English (pp. 118-9). - This chapter closes, as is typical throughout the book, with a formalization of the proposed analyses within Optimality Theory.

    Chapter 4 (pp. 135-191) defends the already widely accepted view that Old English vowel alliteration was based on glottal stop epenthesis. Convincingly, Minkova shows that Old English required a filled stress syllable onset. This assumption is nicely argued through elision in Old English verse (pp. 145-150), the clear observance of morphological boundaries (pp. 150-160), and, though less convincingly, through irregular spellings (pp. 160-5). In the last mentioned case, the use of occasional inorganic <h> in prevocalic positions is interpreted as representing examples of glottal stop epenthesis (p. 163). No evidence for such spelling conventions in other older Germanic languages, e.g. High and Low German, is shown though. For the thesis to be believable similar non-etymological spellings would be desirable. Nevertheless, enough other compelling reasons are given which make glottal stop epenthesis the most likely option for Old English. However, in contrast to, for example, glottal stop epenthesis in Standard German, it is assumed that the Anglo-Saxons were aware of this feature, and with it created alliterative euphony. Minkova goes on to make a convincing argument that such epenthesis had become optional by Middle English. It is shown how the reduction of glottal stop epenthesis led to a blurring of morphological boundaries; also the new tendency to alliterate on like vowels might point towards loss of glottal stop as the main feature behind such vowel alliterations. Finally, it is suggested that loss of glottal stop epentheses in Middle English may be due to Anglo-Norman influence (pp. 166-167). But as the bulk of the population remained English speaking in the centuries following the conquest, it seems far from certain whether borrowing of Romance words and English-Anglo-Norman bilingualism among socially prestigious speakers could have ousted this subphonemic feature.

    Chapter 5 (pp. 192-237) deals with cluster alliteration in Old English. It is asked why only /sp-, st-, sk-/ clusters were treated as units in alliteration by the scops. An approach based on cluster cohesiveness is put forward. It is pointed out that the special nature of these clusters can also be observed in modern languages as seen, for instance, in currently attested metatheses, speech errors, and spoonerisms (p. 208). However, Minkova has also been able to draw on perceptually-based research into consonant clusters currently being carried out at UCLA (cf. in particular Fleischhacker 2001). It has been noted, for example, that s + stop clusters are particularly resistant to separation (p. 226), containing no so-called 'perceptual-breaks' as found in C + resonant clusters, e.g. /pl-, tr-, br-/, which are sometimes liable to separation in Middle English verse. This thesis of cluster cohesion is actively pursued in chapters 6 and 7 and appears to represent a promising line of research. It was a disappointment, however, not to find some brief references to earlier studies into cohesion, e.g. Bell (1979). Indeed quite a few insights found in chapter 5 and subsequent chapters of the book were already brought forward in one form or another in Vennemann (1988). For instance, the advantages of a cohesion-based approach to s + stop clusters as opposed to treating 's' as an extrametrical prependix, see Vennemann (1988).

    A further important discussion point of chapter 3 involves the question of when and how the Old English cluster sk- <sc> was palatalized and assibilated in Old English. Using alliterative evidence, e.g. usage in Aelfric's alliterative prose (p. 200), Minkova concludes (p. 201) that the loss of the bisegmental status of sk- cannot be established with certainty for all dialects until after c. 1100.

    Chapter 6 (pp. 238-310) looks at the distribution of cluster alliterations in Middle English. For this the Middle English compositions Lagamon's Brut, Wynnere and Wastoure, The Wars of Alexander, and Piers Plowman are analysed. It is demonstrated how Middle English poets ignored the special status of /sp-, st-, sk-/ cluster alliteration. Minkova observes that poets of fourteenth- century verse could treat all consonant clusters (e.g. /br-, tr-, kl-/) as units (p. 308): ''all clusters were judged to be cohesive, but not equally so.'' To this end Minkova proposes a hierarchy of cohesiveness for English cluster onsets based on her preliminary statistical findings of Middle English alliterations. Clusters /st-, sp-, sk-/ are found most cohesive, followed by the clusters s + sonorant /sn-, sl-, sl-, sw-/, stop + sonorant /pr-, br-, pl-, kw-/ etc., and, lastly, fricative + sonorant /fr-, fl-/ etc. (p. 305). In Minkova's view, the distribution of alliterative matching in verse (p. 308) ''supports the idea of a universal phonological hierarchy of perceptual cohesiveness in onset clusters.''

    Finally, in chapter 7 (pp. 311-370) the developments of historically unstable clusters /kn-, gn-, hn-, hr-, hl-, hw-, wl-, wr-/ are treated. The four texts studied in chapter 6 again provide the main basis for investigation, but are backed up with reference to the Middle English Dictionary and unpublished Linguistic Atlas of Early Medieval English material. It is argued using verse and scribal evidence that there is no compelling evidence to suggest that gn-clusters were simplified prior to kn- clusters, similarly that wl-clusters were simplified before wr-clusters (note modern spelling, however), or that hn-, hl-, and hr-clusters were simplified before hw- clusters (p. 369). As Minkova notes (p. 369), this stands in contrast to accounts based on consonantal strength relations, in particular Lutz (1991).

    Special attention is drawn to the developments of the cluster hw- in the various dialects (pp. 349-365). Minkova argues that the alliterative evidence (e.g. in Lagamon's Brut, Caligula MS) suggests its simplification to w- in certain southern dialects as early as the 12th century (pp. 313-316, 369). Its survival, especially in the North of England - as exemplified, for instance, in The Wars of Alexander (here alliteration on hw:hw or hw:qu is the rule rather than hw:w) - is accounted for by a differing development of the cluster, most importantly a frication of hw- to xw- (with phonetic variations) in especially Northumbrian dialects. This idea is an old one; long assumed, for example, on account of scribal forms such as <qu(h)> in Late Middle English and by the longer preservation of hw- in traditional northern English dialects. In a nutshell, Minkova proposes that there were several allophonic variations of <hw-> in Early Old English, and that by Late Old English a fricative pronunciation became most common in the North and a pre- aspirated variant in the South (pp. 355-357). This would explain the Middle and Modern English reflexes, but does not really explain why the regional variance occurred. One problem with this account is that it leaves out of picture the interrelated development of English kw- which, surprisingly, is not also classed as an unstable cluster.

    In Northumbrian dialects it was unstable as Middle English verse evidence as well as Early Modern English and even contemporary evidence shows (see Laker 2002 with references). The frication of hw- and the spirantization of kw- resulted in the merger of the two forms (kw-, hw- > xw-) and ought to be seen as a single unitary development in Northumbrian England, cf. Ekwall (1922), Orton (1933), Kristensson (1967). Any explanation for the development of Old English <hw-> must therefore look at all interrelated developments. A new proposal along these lines with reference to Brittonic language contact is suggested in Laker (2002).


    Using alliterative verse, Minkova provides many interesting new analyses of the actuation and chronology of Early English sound changes. Some proposals seem more convincing than others, and some may be in need of further refinement. Nevertheless, by the end of the book Minkova succeeds in demonstrating why verse evidence should be figuring more prominently in discussions on English historical phonology. Apart from phonology, however, the book also provides a mine of insight and information on the English and Germanic poetic tradition in general. It therefore makes essential reading for anyone interested in these fields of study.


    Bell, Alan. 1979. The syllable as constituent versus organizational unit. In Paul R. Clyne et al. (eds.) The elements: a parassession on linguistic units and levels, Chicago: Chicago Linguistic Society, 11-20.

    Bliss, Alastair J. 1967. The metre of ''Beowulf.'' Revised edn. Oxford: Blackwell.

    Cable, Thomas. 1974. The meter and melody of Beowulf, Urbana: University of Illinois Press.

    Ekwall, Eilert. 1922. The place-names of Lancashire, Manchester: Manchester University Press.

    Fleischhacker, Heidi. 2001. Onset cluster behavior in alliteration and reduplication. Phonology Seminar Handout, February 6, 2001, University of California, Los Angeles.

    Hogg, Richard. 1992. A grammar of Old English, vol. 1 Phonology, Oxford: Blackwell.

    Keating, Patricia and Aditi Lahiri. 1993. Fronted velars, palatalized velars, and palatals. Phonetica 50, 73-101.

    Kristensson, Gillis. 1967. A survey of Middle English dialects 1290-1350: the six northern counties and Lincolnshire, Lund: Gleerup.

    Laker, Stephen. 2002. An explanation for the changes kw-, hw- > xw- in the English dialects. In Markku Filppula et al. (eds.) The Celtic roots of English, (= Studies in languages 37), Joensuu: Joensuu University Press, 183-198.

    Lutz, Angelika. 1991. Phonotaktisch gesteuerte Konsonantenver�nderungen in der Geschichte des Englischen, T�bingen: Niemeyer.

    Nielsen, Hans F. 1985. Old English and the continental Germanic languages: a survey of morphological and phonological interrelations, 2nd ed., Innsbrucker Beitr�ge zur Sprachwissenschaft, Bd. 33.

    Orton, Harold. 1933. The phonology of a south Durham dialect: descriptive, historical, and comparative, London: Tr�bner.

    Schuhmacher, Karl. 1914. Studien �ber den Stabreim in der mittelenglischen Alliterationsdichtung (= Bonner Studien zur englischen Philologie 11), Bonn: Hannstein.

    Stockwell, Robert and Donka Minkova. 1997. Prosody. In Robert E. Bjork et al. (eds.) A Beowulf Handbook, Lincoln: University of Nebraska Press, 55-85.

    Vennemann, Theo. 1988. The rule dependence of syllable structure. In Caroline Duncan-Rose et al. (eds.) On language: rhetorica, phonlogica, syntactica: a festschrift for Robert P. Stockwell from his friends and colleagues, London and New York: Routledge, 257-283.

    Vennemann, Theo. 1995. Der Zusamenbruch der Quantit�t im Sp�tmittelalter und sein Einfluss auf die Metrik. In Hans Fix (ed.) Quantit�tsproblematik und Metrik: Greifswalder Symposion zur germanischen Grammatik (= Amsterdamer Beitr�ge zur �lteren Germanistik, 42), Amsterdam: Rodopi, 185-223.


    Stephen Laker is a PhD candidate at the University of Leiden Centre for Linguistics, currently researching language variation and change in Modern and Medieval English.