[Lingtyp] Do experimental and typological studies predict each other?

Heath Jeffrey schweinehaxen at hotmail.com
Mon Jan 15 15:57:13 UTC 2018

If you look at stem-internal diminutivization processes (i.e. vocalic and/or consonant mutation) rather than adjectives meaning 'small' or nouns like 'child', you get stronger sound-symbolic correlations, though the number of languages is smaller. Those adjectives and nouns that are consistent with sound symbolism in typological samples may have originally been subject to such stem-internal processes, now perhaps frozen, but not all such adjectives and nouns have any diminutive phonetic features.

Australian languages are not rich in stem-internal ablaut, so far as I know, but there are embryonic outbreaks of it, such as Nunggubuyu (aka Wubuy) wirig 'small' and its diminutive wiɲig 'little, teeny', their irregular plurals wuraayuŋ and wuɲaaɲuŋ, and two other lexical pairs with r/ɲ alternation. wiɲig is likely < *gaɲaʔ 'small'. It not only shows an etymologically incorrect (but sound-symbolically correct) shift to i-vowels, it also has its w harden to b instead of etymologically correct g after a nasal, hence -biɲig rather than *-giɲig. Even the less intensively diminutive wirig 'small' does this (-birig after nasal), though it too probably originated with initial *g (cf. -girikiriɲ in a neighboring language). Not only do these developments resonate with vocalic sound symbolism (high F2 = small, cute, endearing) and consonantal sound symbolism (palatalization = high F2 = ...), they also resonate with an interesting tendency to intensify the high F2 by juxtaposing a rounded vowel or semivowel (all formants low), exactly as in teen(s)y-ween(s)y and wee, Arabic diminutives beginning Cwi... or Cwiyy... with nonlexical w, and with inverted syllabicity English cute, choo-choo 'train', and so forth. It is in these embryonic developments, and not in sweeping typological surveys based on operational definitions of target forms that are loosened to be applicable to all/most languages, that the power of sound symbolism is manifested.

Similarly, in sociophonetics, sound symbolism is manifested most clearly in relatively abrupt sharp movements of individual vowels in the æ → e → i direction (generally favored by urban/middle-class women at least in developed societies), or less often in the direction of ɔ (mostly men), rather than in crossdialectal comparisons of total vowel systems which reflect subsequent recalibrations of vowel positions (Current Anthropology 39: 421ff.).

This is one example among many why I prefer close synchronic and historical study among closely related languages to typology as practiced since Greenberg, even when the ultimate goal is crosslinguistic.

From: Lingtyp <lingtyp-bounces at listserv.linguistlist.org> on behalf of Claire Bowern <clairebowern at gmail.com>
Sent: Monday, January 15, 2018 7:46:39 AM
To: JOO Ian
Cc: lingtyp at listserv.linguistlist.org
Subject: Re: [Lingtyp] Do experimental and typological studies predict each other?

For some information about this in an Australian context, see Haynie, Bowern, and LaPalombara (2014): http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0092852<https://nam03.safelinks.protection.outlook.com/?url=http%3A%2F%2Fjournals.plos.org%2Fplosone%2Farticle%3Fid%3D10.1371%2Fjournal.pone.0092852&data=02%7C01%7C%7Cb8cb7cef1d6047a97ba808d55c17205a%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636516176750098890&sdata=Bcl3oViVNpMeWW%2FvLqN9qx4B2XtaJ%2BmwNfacpDxi12E%3D&reserved=0>.. We found some support but not in every language we looked at.

On Sat, Jan 13, 2018 at 7:04 AM, JOO Ian <il.y.en.a at outlook.com<mailto:il.y.en.a at outlook.com>> wrote:

Dear all,

Many experimental studies show that people tend to associate high front vowels with small size and low back vowels with large size (e. g. Shinohara and Kawahara 2010).

Bauer’s (1996) study, on the other hand, show that diminutive and augmentative affixes are not correlated with specific vowels, contrary to what one would expect based on experimental studies.

This leads me to think that experimental and typological studies do not always predict each other. That is, the correlations demonstrated by experiments are not necessarily statistically visible in natural languages, what is statistically significant in natural languages may not be demonstrable through experiments.

What do you think about the predictability between experimental and typological studies? Can you think of any example where there is no predictability, like the case of Bauer (1996)?


Ian Joo



Bauer, Laurie. “No Phonetic Iconicity in Evaluative Morphology.” Studia Linguistica, vol. 50, no. 2, 1996, pp. 189–206.

Shinohara, Kazuko, and Shigeto Kawahara. “A Cross-Linguistic Study of Sound Symbolism: The Images of Size.” Annual Meeting of the Berkeley Linguistics Society, vol. 36, 2010, pp. 396–410.

Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20180115/a8117af0/attachment.htm>

More information about the Lingtyp mailing list