Is the Analytic Unit ‘Syllable’ a Crosslinguistic Universal?

Astrophil
10 min readMay 2, 2022

--

Prosodic hierarchical structure of ‘pretty butterfly’
  1. Introduction

In discourses, such as Hyman (2011) and Labrune (2012), the generally acknowledged universality of syllable has been challenged. As with many phonological universals’, the enormously thorny testification, not to mention falsification, of concept of syllable complicates the discussion. Is it, however, genuinely necessary to identify a phonological entity as a universal? This essay summarises and re-examines arguments for and against syllable universality in an attempt to echo the view that the concept of syllable is highly theory-dependent ­– while useful as an analytic unit, it is not a requisite — and redirect the locus of research on phonological theory — if a notion is beneficial to analysis, instrumentalise it; if not, why do so?

2. Background

2.1. The syllable

The syllable, as a critical analytic phonological unit, has been of critical interest of a large body of literature on prosody. Many phonologists oftentimes regard it as a cross-linguistic universal in phonology. Nevertheless, until today, an unambiguous, uniform linguistic definition remains absent. The majority of phonological definitions are language-specific. Generally, the syllable is recognised as the structural unit that provides melodic organisation of segment strings (Blevins, 1996), as well as a representational device for phonotactics, which contains restrictions on segment distribution (Zec, 2006).

The syllable is, moreover, a notion of crucial analytic and descriptive function in the sense that, subjective to duplicative processes, it not only serves as the domain of the occurrence of phonological alternations, stress assignment, and tonal specification, but also accounts for a multitude of omnipresent cross-linguistic similarities among permissible segment sequences. In addition, it is considered an abstract phonological constituent that neither has direct phonetic, phonological, and articulatory correlates (Ladefoged & Maddieson, 1996) that are cross-linguistically applicable nor crosslinguistic indicators that pinpoint the boundaries of a syllable.

2.2. Controversy on syllable universality

Since Noam Chomsky’s postulate of Universal Grammar (UG), the quest for crosslinguistic grammatical universals has become a mainstream trend in phonological research. However, as breaches of UG such as Pirahã’s lack of recursion (Everett, 2012) gradually surface, one cannot but ponder whether universality truly exists. Should the pursuit of absolute universals continue to dominate discussions on phonological theory?

Like a large number of phonological universals, the syllable is exceedingly difficult to testify and thus to disprove. No concrete evidence for the existence of the syllable has been presented thus far (Labrune, 2012). The difficulty is primarily due to the phonological abstract nature of the syllable, that is, it is conceivable for syllables to be absent or not exhibit any obvious evidence for their existence on the surface yet present underlyingly in languages, rendering verification empirically impossible (Hyman, 2011). In addition, prior to the development of non-linear phonology in the 1970s and the resurgence of interest in examining prosodic phenomena, the notion of syllable is, in fact, absent in framework of generative phonology proposed by Chomsky and Halle (1968) (Labrune, 2012); it may virtually be unnecessary to assume the syllable as an absolute universal.

The last decade has seen challenges on syllable universality such as those made by Hyman (2011) and Labrune (2012) using examples of Gokana and Tokyo Japanese respectively. On this syllable-less end of the spectrum, after close scrutiny, Hyman (2011) has only discovered limited, insignificant and ambiguous evidence for the presence of the syllable, i.e., the mono- or bi-syllabicity of feet, which a prosodic stem can only contain one, with either syllable being heavy or light, in Gokana; thus, he argues that while syllables are extensively activated in certain languages, they are less activated in others. Meanwhile, Labrune (2012) has presented evidence from, for instance, speech errors and absence of onset optimisation, pointing to that heavy syllables are irrelevant in Tokyo Japanese phonology and that morae are, in turn, the basic prosodic unit of the language.

A more extreme view shared by some scholars is that all tiers of the prosodic hierarchy are entirely optional, with syllable serving nothing more than a possible constituent (Labrune, 2012). Ito & Mester (2003) advocates a reinterpretation of the prosodic hierarchy as a universal inventory of choices, in lieu of an absolute universal. What goes further are discourses, such as Hyman (2011) and Labrune (2012), that attempts to formalise the status of mora as no longer a subpart of the syllable, but the genuine independent prosodic constituent, whose operation need not the licensing of syllables, in order to accommodate a broader variety of different linguistic realities. Such a conception lends credence to the property-driven approach to phonological typology and universals advocated in Hyman (2011), which focuses on languages’ systemisation of cross-linguistically available phonetic substance, unlike mainstream approaches, which aim to categorise languages and hunt for the gamut of phonological universals.

3. Of universality

3.1. What is universality?

The preceding section has presented an overview of the controversy on syllable universality. Nonetheless, discussions would be nugatory without defining the fundamental criterion that a phonological entity must fulfil in order to be regarded as a universal. Adapting notions from Evans & Levinson (2009) and Hyman (2011), this section enumerates two contrasting possibilities:

(1)

a. Marginal exhibition in phonology of all languages and

b. As a focal organisational principle of all languages.

In the context of discussing syllable universality, what (1a) denotes is that as long as the syllable is marginally exhibited in the phonology of all languages, they are considered universal; on the contrary, (1b) denotes that the syllable must be employed by all languages as the focal organisational principle in order to be considered universal.

Should (1a) is the requisite, the case of Gokana in Hyman (2011) mentioned in section 3 might be consistent with syllable universality. The evidence is that assuming that the maximal prosodic stem structure CVV(C)VV being two syllables, each of which has the maximum of two morae, Hyman (2011) reports that the impermissible structures, *CVCVVV and *CVVVCV, are thus accounted for, but with ambiguity stemming from its exclusive pertinence to verbs.

However, of critical interest is whether marginal involvement of this sort in phonology is in any way theoretically significant. Are these attempts on eliminating language-specificity purely for the sake of a higher restrictiveness of linguistic theory on the prosodic levels (Kawahara, 2016) and/or conformity to UG? Is it possible for a language to function entirely without a property that is marginally exhibited in all languages (Hyman, 2011)? Should there be not much of an actual benefit to phonological discussions, perhaps (1b) may be a more appropriate criterion.

If (1b) is the criterion, then Gokana would apparently be a counterexample of syllable universality in that regardless of the ambiguity of evidence for syllable presence, it is not the syllable that has been employed to organise the phonology; instead, it is the mora (Hyman, 2011). It may be possible that the syllable, similar to recursion, is, after all, a capacity that languages potentially display, in lieu of a universal feature.

The lack of a uniform definition of universality is detrimental to phonological research in a way that pertinent discussions would be pointless if scholars have disparate understanding on universality, and hence, their findings would be inapposite for integration in order to construct a complete picture of phonological realities.

3.2. Impacts of universality

With a brief review of universality in the previous section, it may be appropriate to ponder whether universality has an impact on phonological and/or linguistic research. Hyman (2011) points out that it does not make much of a difference and hardly influences phonology. Henceforth, the emphasis of phonological theory should be on why a multitude of languages collectively refer to the syllable, or in other words, how phonetic substances are varyingly phonologised across languages, rather than the fact that languages have diverse degree of reference to it (Hyman, 2011). Indeed, universality proffers theoretical coherence and demonstrates the mutually shared underlying cognitive and/or biological characteristics of humans (Li, 2014); notwithstanding, as an increasing number of examples contradicting previous universal emerge, we should re-contemplate the necessity of establishing universals. After all, the issue of universality might not, in turn, be so clearcut: some languages activate the syllable more extensively whereas some less (Hyman, 2011).

4. The conventional universal-syllabic framework

4.1. Advantages

Aside from convenience of analytic purposes, Kawahara (2016) enumerates the benefits for the syllabic framework: the assumption that all languages contain syllables (i) provides simplicity for language learners during language-acquisition process and (ii) lends support to a more restrictive linguistic theory that disallows language-specificity in prosody.

4.2. Disadvantages

As pointed out by Kohler (1966), any collisions between syllabic divisions and morphemic divisions, as well as obscurities of underlying phonological structures as a result of the syllable indicate the harmfulness of the notion of syllable. The English language, in which both conditions apply, is an exemplar, c.f., (2).

(2) Vowel + lateral conditions in English:

a. Coolish — [uɫ],

b. Foolish — [ul], and

c. Ballachulish — [u̟l]

In several accents of English, the above vowel + lateral sequences are realised differently (Kohler, 1966). Owing to dark and clear lateral distribution, the syllable division occurs after the lateral in (2a), before the lateral in (2c), but not after the lateral in (2b), which is directly contradictory to the morphology (Kohler, 1966).

5. A syllable-free approach

In recent discussions on syllable universality, syllable-free analyses, e.g., mora- and/or foot-based analyses, that allow for phonological systems to be thoroughly scrutinised without notable descriptive loss in the absence of the concept of syllable, e.g., properties such as phonological rules can be well-analysed with morae, in Gokana (Hyman, 2011), c.f., (3), are available (Labrune, 2012).

(3) Intervocalic realisations of /B D/ in Gokana

a. /B D/ → [v r] / [–cons] __ [–cons]

b. /B D/ → [b l]

Given that /j/ and /w/ are underlyingly absent in Gokana, intervocalic contexts can be described with [–consonantal], as in (3a), without the traditional reference to syllabicity (Hyman, 2011).

Meanwhile, on the contrary, syllables fail to reformulate mora- and/or foot-based accounts, c.f. (4).

(4) Accentuation of /N R/ in Tokyo Japanese

a. Obaasankko /obaRsaNQko/ — ‘child cherished by grandmother’

b. Cheenten /tjeRNteN/ — ‘chain store’

McCawley (1968) and other syllabic analyses assume that solely the first part of a long syllable receives accent, i.e., the accentual peak of the syllable; however, (4a) and (4b) demonstrates the exceptional cases of accentuation of what Labrune (2012) refers to as special morae, namely the moraic nasal (/N/) and the second part of a long vowel (/R/). In those circumstances, it is the special mora following the nucleus, in lieu of the syllabic accentual peak, that receives the accent.

Some such as Selkrik (1986) contend that morae and feet are not universal. Regardless, such an argument might, in fact, serve to solidify the view that syllables are not universal too; given that all three of them are prosodic constituents, no a priori reasons would justify the assumption that syllables are more superior than others (Li, 2014).

5.1. Downside

Despite not resorting to syllabicity, the syllable-free is essentially a recapitulation of the syllable theory in a different terminology, i.e., ‘full’ and ‘deficient’ morae, as Ito and Mester (2015) remark. Even the reanalyses phonological patterns undertaken by proponents of the syllable-free approach such as Labrune (2012) indicate that despite possible without the syllable, they unnecessarily conducted that way (Kawahara, 2016).

In addition, certain arguments of the syllable-free approach, e.g., those of Labrune (2012) pertaining to how the lack of gradual changes between two adjacent vowels signals the absence of syllables, are questioned as a result of lack of quantitative support (Kawahara, 2016); in opposition, such a claim was, in turn, dismissed by Kawahara (2012) using spectrogram and waveform data.

The majority of works on the syllable-free framework resort to mora as a fundamental unit. Nevertheless, as noted by Kawahara (2016), in the example of Tokyo Japanese, morae are admittedly crucial, particularly in segmentation, but so are syllables; the role of syllables has become increasingly palpable via experimentation, as well as assessment of speech behaviour of native speakers.

Besides, abandoning the syllable concept would potentially lead to repercussions for various aspects of phonological theory, including but not limited the loss of what has been outlined in section 5.1.

6. Conclusion

This essay has thus far provided both summaries and examination of arguments from both syllable-free and syllable-universal approaches, albeit necessarily selective due to length limitations. On account of the deviating definitions of the syllable and universality, as well as theoretical assumptions, syllable universality is highly theory-dependent. Despite that the syllable is instrumental as an analytic unit, it is not always obligatory. ­If it is beneficial to phonological analysis, it would be sensible to employ it; if it is not the case, there should be no compelling reasons for phonologists to adopt it. It might perhaps be more consequential to ascertain what outlier systems reveal regarding the general nature of phonological systems (Hyman, 2011).

References

Blevins, J. (1996). The Syllable in Phonological Theory. In Goldsmith, John A, The Handbook of Phonological Theory. Oxford: Blackwell.

Chomsky, N. & Halle, M. (1968). The Sound Pattern of English. New York: Harper & Row.

Evans, N. & Levinson, S. C. (2009). The Myth of Language Universals: Language Diversity and Its Importance for Cognitive Science. Behavioral and Brain Sciences, (32), 429–448.

Everett, D. (2012). Language: The Cultural Tool. London: Profile Books.

Hyman, L. M. (2011). Does Gokana Really Have No Syllables? Or: What’s So Great About Being Universal? Phonology, 28(1), 55–85.

Ito, Junko & Armin Mester (2003). Weak layering and word binarity. In Takeru, H., Masao, O., Toshiyuki, T., & Shin’ichi, T. (eds.). A New Century of Phonology and Phonological Theory: A Festschrift for Professor Shosuke Haraguchi on the Occasion of his Sixtieth Birthday. Tokyo: Kaitakusha. 26–65.

Ito, Junko & Armin Mester (2015). Word Formation and Phonological Processes. In Kubozono, H. (ed.). The Handbook of Japanese Phonetics and Phonology. 363–395.

Kawahara, Shigeto (2012). Review of Labrune (2012b). Phonology 29. 540–548.

Kawahara, S. (2016). Japanese Has Syllables: A Reply to Labrune. Phonology, 33(1), 169–194.

Kohler, K. J. (1966). Is the Syllable a Phonological Universal?. Journal of Linguistics, 2(2), 207–208.

Labrune, L. (2012). Questioning the Universality of the Syllable: Evidence from Japanese. Phonology, 29(1), 113–152.

Ladefoged, P. & Maddieson, I. (1996). The Sounds of the World’s Languages. Oxford, Blackwell.

Li, W. L. D. (2014). Prosodic Structures of Different Japanese Dialects and the Universality of the Syllable [Master of Arts dissertation, University van Amsterdam].

McCawley, James D. (1968). The Phonological Component of a Grammar of Japanese. The Hague & Paris: Mouton.

Selkirk, E. O. (1986). Phonology and Syntax: The Relationship Between Sound and Structure. Massachusetts: MIT Press.

Zec, D. (2007). The Syllable. In Lacy, P. The Cambridge Handbook of Phonology. Cambridge: Cambridge University Press.

--

--

No responses yet