Core Phonology: The Articulatory Interface

  Core Phonology: The articulatory interface

There is widespread agreement that both phonological knowledge [1,2] and spoken linguistic signals [3] manifest putative universals, which, systematically used, enable infinite productivity. However, less agreement prevails on whether the phonological system is combinatorial [4] or blending [5] or in nature. In a combinatorial system (DNA is an example), discrete units are combined recursively to yield infinite forms, while maintaining the individuality of each block. For linguistic forms, optimal transmission generally assumes that some contextual blending of phonemes is necessary [6]. It has been variously claimed that transmissibility thus renders phonology merely a summary requirement imposed by the speech channel [6], while others argue that such purely blending systems cannot account for highly unnatural yet productive patterns that have been attested [7], and that the greedy reductionism of cognitive abilities to motor functions amounts to substance abuse[8]. An alternative view holds that while phonology is indeed algebraically combinatorial [9], it is also subject to functional pressures [10] albeit only selectively [9]. We here report a test of the latter alternative hypothesis.

Two modified “WUG tests” [11] were conducted with 15 native speakers of Malayalam, making use of the Malayalam paradigm of velar palatalization, whereby single-melody velars are palatalized when preceded by [i, e, a], both within and across word boundaries [12]. Two palatalizing suffixes figured in our first experiment ([k:uka] “INF” and [k:ə] “DAT”), along with a non-palatalizing suffix ([kal]) in the second. A male native speaker recorded samples of nonce stems ending in [i, e, a] and the three suffixes, separately. Before testing, participants were given examples of an unrelated stem (e.g., /adhyāpika/ “teacher-FEM”) and suffix (e.g., [-mār] “num-Pl”), and then an instance of a conjugated form (e.g., /adhyāpikamār/ “teachers-FEM”) to explain what conjugation involves. During testing, participants listened to randomly selected pairs of a nonce stem and a suffix with a 5 ms gap in between (e.g., [bati] – 5 ms – [k:uka]), and they were instructed to speak out the conjugated form ([batik:uka]). Their responses were recorded using a Roland UA 25-EX sound card, and the tokens were categorized by three native listeners into palatalized and non-palatalized forms. We also compared the spectral properties of the categorized velars (centroids, intensity and durations of consonantal bursts) and the preceding and following vowels (F2).

Confirming the native listeners’ intuitions, the acoustic measures revealed a systematic difference between palatalized and unpalatalized forms. Mean F2 values of vowels were highest for [i] and lowest for [a] ([i] > [e] > [a]), while palatalized forms of each vowel demonstrated higher F2, indicating frontedness, relative to unpalatalized forms. Likewise, palatalized velars displayed higher mean centroid, intensity and duration values than did unpalatalized velars. In Experiment 1 the palatalization pattern revealed a main effect of vowels, F(1,14) = 7.56, p < 0.01, and a lower but significant suffix effect, F(1,14) = 6.52, p = 0.023, but no interactions. Planned contrasts revealed that vowels [i] and [e] were better triggers of palatalized articulation,than [a] , p < 0.01, but no difference was observed between [i] and [e], p > 0.05. The two suffixes differed slightly in their rate of palatalizing the forms, p < 0.05. Experiment 2, likewise, produced main effects of vowels, F(1,14) = 10.38, p < 0.01, and suffixes, F(1,14) = 1747.79, p < 0.01, but no interactions. Planned contrasts on the vowels conformed to the findings reported in the previous experiment, with the exception that all vowels failed to exhibit triggering with the additional unlicensed suffix [kal] included in Experiment 2; for suffixes, we found the two palatalizing suffixes to be significantly higher palatalization triggers than the unlicensed [kal], p<0.01, but the two licensed suffixes did not differ. These findings conform overall with the suffix-licensing proposal for Malayalam by Mohanan [12], with three novel empirical findings: (1) while the phonetic implicational laws of palatalization [13] prescribe a gradient scale for vowels triggering palatalization ([i > e > a]), in our data only the most typologically unnatural vowel [a] differed gradiently from [i, e], which did not differ from each other; (2) the unnatural vowel [a] nevertheless shows highly productive palatalization with the licensed suffixes, while all vowels fail to trigger it with the unlicensed suffix; and (3) a gradient suffix effect is observed in the triggering rates of the licensed suffixes in the first experiment, whereas in Exp. 2 both the licensed suffixes differ significantly from the unlicensed one but not from each other.


We interpret these findings as reflecting the double duty of the phonological system: ensuring infinite productivity, including generalizability with unattested novel forms, while maintaining optimal transmissibility. The proposal of The Computational Theory of Mind [14] is that infinite arrays of contrasts can be produced by positing discrete algebraic variables grouped into equivalence classes, and prescribing rules that operate upon these classes (e.g., IDENTITY restrictions: *C1__C2; C1, C2, C3...Cn ε [C]). Thus in our data an equivalence class of triggering vowels ([i, e, a]) triggers a palatalizing rule with an equivalence class of licensed suffixes ([k:uka, k:ə]), while rejecting suffixes that are treated distinctly from members of this equivalence class (e.g., [kal]). Such computations in phonology may, however, be sensitive to functional phonetic concerns (note the lower triggering rates with [a], as well as the suffix effects in Exp.1), though such concerns are readily ignored if they conflict with the grammar (note the naturalness concerns violated by [kal], and the change in suffix effects in Exp.2). Similar findings are reported elsewhere from a perceptual forced-choice experiment involving the same paradigm [15], which were ascribed to effects stemming from the externalization of language (a digital, computational-representational system: CR) by the sensory-motor system (SM, an analog system). In the perceptual experiments no suffix effects at all were observed. The presence of such effects in the task reported here is, we claim, due to thecompetence-performance difference. Articulation involves greater motor control, and is subject to physical constraints such as inertia. Therefore, articulatory tasks are more likely to manifest demands of the SM system that are not always perfectly suited for the CR system. However, our data clearly illustrate that while phonological processes are, in fact, affected by channel concerns, these are attended to selectively in a manner that does not override grammatical requirements. In light of the distinct but converging findings from the two experiments, we conclude that phonology manifests algebraically combinatorial rules that are selectively subject to functional optimization in order to attain discrete infinity while maintaining transmissibility.


1. Berent, I., Shimron, J., & Vaknin, V. (2001). Phonological constraints on reading: Evidence from the Obligatory Contour Principle.

Journal of Memory and Language, 4(4), 644–665.

2. Chomsky, N. (1959). On certain formal properties of grammars. Information and Control, 2(2), 137–167.

3. Liberman, A. M., Cooper, F. S., Shankweiler, D. P., & Studdert-Kennedy, M. (1967). Perception of the speech code. Psychological Review, 74(6), 431.

4. Berent, I., Everett, D. L., & Shimron, J. (2001). Do phonological representations specify variables? Evidence from the Obligatory Contour Principle. Cognitive Psychology, 42(1), 1–60.

5. Abler., W.L.(1989). On the particulate principle of self-diversifying systems. Journal of Socials and Biological Systems, 12, 1-13.

6. Pierrehumbert, J. (1990). Phonological and phonetic representation. Journal of Phonetics, 18(3), 375–394.

7. Berent, I., & Lennertz, T. (2010). Universal constraints on the sound structure of language: Phonological or acoustic? Journal of Experimental Psychology: Human Perception and Performance, 36(1), 212.

8. Hale, M., & Reiss, C. (2000). “Substance Abuse” and “Dysfunctionalism”: Current Trends in Phonology. Linguistic Inquiry, 31(1), 157–169.

9. Berent, I. (2013). The phonological mind. Cambridge: Cambridge university press.

10. Hayes, B. P. (1999). Phonetically driven phonology. Functionalism and Formalism in Linguistics, 1, 243–285.

11.Berko, J. (1958). The child's learning of English morphology. Word, 14(2-3), 150-177.

12. Mohanan, K. P., & Mohanan, T. (1984). Lexical phonology of the consonant system in Malayalam. Linguistic Inquiry, 575-602.

13. Wilson, C. (2006). Learning Phonology With Substantive Bias: An Experimental and Computational Study of Velar Palatalization. Cognitive Science, 30(5), 945–982.

14. Fodor, J. A. (1983). The modularity of mind: An essay on faculty psychology. MIT press.

15. Mandal, S.S, Best, C.T., Shaw, J.& Cutler, A. (forthcoming). Do rules rule? Evidence from Malayalam velar palatalization.