Re: Most common consonant cluster types cross-linguistically

On Sat, 16 Aug 2008 08:47:30 -0400, Jim Henry <jimhenry1973@...>
wrote:
>Is there any comparative data available on what are the most
>common kinds of consonant cluster across different languages?
>E.g., my impression is that nasal + nasal and plosive + plosive
>onset clusters as in Greek (mnemo, ptera, etc) are rarer than
>clusters that mix different manners of articulation, like fricative + plosive
>or plosive + fricative; but how much has this been quantified?
>A Google search for "most common consonant clusters" didn't
>turn up anything precisely relevant.  The closest fit
>was this CONLANG message from 2002 by Christophe Grandsire,
>
>http://archives.conlang.info/vhu/wilso/phinthofian.html
>
>helpful, but not as quantitative as I'd like.
>
>--
>Jim Henry
This won't answer your question because you wanted to know what the
empirical data are, and I don't have that yet.

This is just my impression.

The most common broad type of consonant-clusters (I guess) are homorganic
(same place-of-articulation) with different manners-of-articulation.

If you count affricates as clusters, then homorganic stop+fricative (/ts/) and
homorganic fricative+stop (/st/) are almost surely the most common.

Homorganic nasal+stop (/nd/, /mb/, /Ng/, /nt/, /mp/, /Nk/) are probably also
rather common.

The second most common broad type of consonant-clusters (I guess) are
probably those with the same manner-of-articulation but different places-of-
articulation.  /bd/, /db/, /zv/, /vz/, /mn/, /nm/, etc.  In English these tend to
be at syllable-boundaries rather than in the onset or the coda of the same
syllable; maybe that's the case cross-linguistically as well, but I'd imagine
word-internal "sandhi"-like mutation might lead to allomorphy causing these to
crop up frequently even in other languages.

Finally, there is pre-aspiration, post-aspiration, pre-palatalization, post-
palatalization, pre-labialization, post-labialization, and "ejective".  As you know
these are rather common.  TTBOMK, and therefore IMO, these are the most
common types of consonant-clusters that are not homorganic and have
different manners-of-articulation.

A smaller set would have the same P.o.A. and the same M.o.A. but different
voicing; /dt/, /td/, /sz/, /zs/, /gk/, /kg/, /bp/, /bp/, /DT/, /TD/, and so on.  I
would be surprised if these were more common than any of the above; but
surely they do happen -etically if not -emicallly.