LINGUIST List 11.523

Fri Mar 10 2000

Disc: Underlying Shwa?

Editor for this issue: Karen Milligan <>


  1. Yongsung Lee, Re: 11.511, Disc: Underlying Shwa?
  2. Geoffrey S. Nathan, Underlying schwas

Message 1: Re: 11.511, Disc: Underlying Shwa?

Date: Thu, 09 Mar 2000 16:45:27 +0900
From: Yongsung Lee <>
Subject: Re: 11.511, Disc: Underlying Shwa?

We can also think of the phonological aspects of English to
answer the question about underlying schwas. If the surface
schwas alternate with other vowels on the surface, we surely
posit a full vowel in the underlying representation and
resort to vowel reduction to explain the schwas. But if
there is no surface alternation as shown in the typical
example of SPE, effVrt, amVzon and others, (cf. SPE, 37,
footnote 27) we find no evidence of positing any full vowel
in the underlying forms. The weak vowels in these cases can
be viewed as underlying schwas.

Further consider the following contrast: (Data from Lee
(1999), IUPWL)
delta - deltaic
algebra - algebraic
zebra - zebraic
aorta - aortic
vanilla - vanillic
apnea - apneic (cf. delta - deltic, deltaic)

All the words in (a) and (b) end in a surface schwa before
{-ic} suffixation. But they differ after the suffixation.
In (a) the final schwa alternates with a full vowel, while
it deletes in (b). One plausible way to explain the
difference is to posit underlying schwas for the words in
(b). With the assumption that a schwa cannot be stressed,
and that {-ic} must be added to a stressed syllable, the
underlying schwa has to delete in (b) in {-ic} suffixation.
The variation shown in /delta + ic/ indicates that the
output seems to depend on how the speaker treats the
word-final vowel. This line of thought may support the
presence of underlying schwas.

Another less strong evidence is also found in m-final stems
in {-atic} suffixation:
film - filmic
rhythm - rhythmic
atom - atomic
phoneme - phonemic
problem -problemATic
symptom - symptomATic
idiom - idiomATic (cf. system - systemic,

Again the choice between {-ic} and {-ATic} seems to be
dependent on the nature of the stem final vowel. Extending
the idea from (a) and (b), we find that AT is inserted
before the suffix, so that the suffix can be added after a
stressed syllable. The data are not plenty, but there are
certain parts of English phonology (such as stress as
discussed by Prof. Fidelholz that renders phonological
support to the underlying schwas.

- ------------------------------------------------
Yongsung Lee
Professor, English Department
Pusan University of Foreign Studies.
- ------------------------------------------------
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Underlying schwas

Date: Fri, 10 Mar 2000 08:12:44 -0600
From: Geoffrey S. Nathan <>
Subject: Underlying schwas

While I entered this conversation somewhat late, and I apologize if this 
issue has already been discussed at an early stage, I feel the necessity to 
inject an additional complicating factor in a discussion that includes 
words like 'cactusia' and so on. This is my strong belief that for (at 
least) English speakers, underlying forms (lexical representations, 
phonemic storage--choose your flavor) are normally not phonological at all, 
but visual/orthographic. And since there's no spelling for schwa, stress 
shifts will always show up as some unreduced vowel or other. Think about 
it: we hyperliterate academic folks probably learn a majority of our 
vocabulary first through our eyes, not our ears, and I believe we store a 
large amount of our vocabulary that way. I wrote a paper on this twenty 
years ago (CLS Parasession, 1979), and I continue to collect evidence, such 
as tip of the tongue phenomena that indicate that primary storage is 
orthographic. My favorite, not surprisingly, is my name. The number of 
times that people call me 'Greg' or 'Glenn' as we pass in the hallway is 
only explicable if the top sort handle in their mental lexicon is the 
letter 'g', not the phoneme /dZ/. I know we teach our students that 
orthography is just a pale reflection of 'real' phonology, but I think the 
psychological reality of orthographic storage is much stronger than we 
would like to believe, and I think it makes the question of underlying 
schwas much trickier to sort out, and citing made up examples based on 
existing words, or anything based on visually presented data is terminally 
contaminated by our five-vowel orthography.

Geoff (with a dZ)

Geoffrey S. Nathan
Department of Linguistics
Southern Illinois University at Carbondale
Carbondale, IL, 62901-4517
Phone: (618) 453-3421 (Office) / FAX (618) 453-6527
 (618) 549-0106 (Home)
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue