LINGUIST List 12.199

Thu Jan 25 2001

Disc: Origins of Human Language

Editor for this issue: Karen Milligan <>


  1. Zylogy, Origins of human language

Message 1: Origins of human language

Date: Wed, 24 Jan 2001 23:10:11 EST
From: Zylogy <>
Subject: Origins of human language

Several months ago a new type of story about the origins of human language 
emerged after conversations during and after the Language Origins Society 
meeting in New Brunswick, NJ. Introduced on LINGUIST later during a 
different discussion (Does "Language" Mean "Human Language"? in early 
November 2000), it needs critical comment. 

Call it the "signal inversion hypothesis". Basic bits and pieces are as 
follows: There is a gross inverse correlation between the length of a "word" 
and the number of phonemes in a language- I didn't make this up, and its 
something informed linguists should be aware of. Generally, from a 
Jakobsonian perspective, the more phonemes, the more features. Many of these 
extra features can be shown to derive from other phonemes (tones on vowels, 
color on consonants, etc.) and lend higher markedness to the segments they 
occupy. There may be a finite maximal feature set for human languages. One 
may perhaps speak of "feature density"? Click languages (such as !kung) have 
the highest number of phonemes.

If the phonology of a click language is so rich with highly marked segments, 
then the language cannot be considered "primitive" in any sense except 
historically. And one often sees the evolution of such complexity with 
monosyllabification- even morphological marks may create new lexical items 
(as in Tibeto-Burman). The overall impression is that more and more can be 
stuffed into a smaller space, within limits.

The signal inversion hypothesis extrapolates feature density until even 
"phonemes" can be considered as entire utterances some time in the distant 
ancestry of man. With just 16 binary feature pairs, with one choice always 
instantiated for each pair, will generate well over 50000 separate units. 
This would of course require the utilization of more than formants 0,1,2,3, 
but that's not necessarily a problem. The big questions are whether a broader 
band approach is feasible, both in the hearing and in the modulation of 

According to Eugene Morton's Motivation Structure theory, different regions 
of an organism's acoustic spectral range appear to be devoted to different 
emotional states, as are differences in tonality versus noise, etc. I wonder 
whether there is a part/whole calculus going on here with application 
dependent upon context. Is there far more information content in an animal's 
acoustic signal than we thought?

If we focus only on time-distributed combinatorics (the sin of 
syntactically-obsessed linguistics for some time now), then we cut ourselves 
off from more holistic signalling possibilities. Similar distaste has 
marginalized the study of ideophones and their formulaic structure for 

Finally, it may be that the hypothesized inverted structure could have much 
in common with polysynthetic predicate structure- a nice zipped up recipe 
giving one just the relevant facts, fast.

The hypothesis is still a-brewin', so all pertinent criticism welcome, of 
whatever stamp. Thanks (I hope).

Best regards to all,
Jess Tauber
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue