Re: Semantic content (was: Ural-Altaic. A fly)

Joachim Pense wrote:

> > We distinguish between form and function. Notions of plurality -- or
> > tense, or comparison, etc. -- are marked by grammatical morphemes, but
> > the grammatical morphemes don't bear meaning in themselves; they cannot
> > exist in themselves. They enter into paradigms.
>
> But for example "pre-" or "un-" cannot stand alone either. They are not
> grammatical morphemes in English (as they do not occur in the complete list
> you gave).
>
> So tell my if the "un-" in "undo", "unload", "unclear", "unhappy" bears a
> meaning "in itself"? And what makes it different from the "-s" in sisters
> or the "-ed" in "killed"?

This dissertation abstract was just published on LINGUIST List (BTW,
what country is "University of Ulster" in?). At the very least, it
indicates that "grammatical morpheme" is an identifiable category, with
psychological reality:

Institution: University of Ulster
Program: Faculty of Social Sciences
Dissertation Status: Completed
Degree Date: 2004

Author: Maria McGuckian

Dissertation Title: The Grammatical Morpheme Deficit in Children with
Hearing
Impairment, Children with Down's Syndrome and Children with
Specific Language Impairment

Linguistic Field(s): Language Acquisition
Morphology
Psycholinguistics


Dissertation Director(s):
Alison Henry

Dissertation Abstract:

Grammatical morphemes have been reported to be problematic for children
whose language learning problems appear to derive from very different
origins, including specific language impairment (SLI) (e.g. Leonard,
Eyer,
Bedore & Grela, 1997), Down's syndrome (DS) (e.g. Chapman, 1995) and
hearing impairment (HI) (e.g. Norbury, Bishop & Briscoe, 2001). This
thesis
was designed to establish whether the nature of the grammatical morpheme
difficulty is similar or different across (English-speaking) groups of
children who present with moderate-HI, DS and SLI. It was hypothesized
that
the grammatical morpheme deficit would be more similar than different
across these distinct clinical groups and that any obvious differences
between the groups would be largely quantitative in nature.

A wealth of production data, including elicited as well as spontaneous
speech data, was collected and analysed so that across-group comparisons
could be made for rates of production of a range of grammatical
morphemes
(11 different forms - tense and non-tense morphemes), for types and
proportions of grammatical morpheme errors and for order of grammatical
morpheme accuracy.

In addressing the key hypothesis of the research there was first
consideration of the nature of the grammatical morpheme deficit for each
of
the three clinical groups under study when compared to mean length of
utterance (MLU)-matched typically developing children. Following that,
explicit comparisons were made between the different clinical groups for
patterns and rates of grammatical morpheme use.

In the main, the findings supported the key hypothesis showing many
similarities between the various clinical groups for grammatical
morpheme
use. However, some intriguing differences also emerged.

The findings showed that the children with HI were sensitive to
frequency
of elements in input; perhaps until they reach a necessary threshold for
spoken language intake. In contrast, the children with DS and SLI did
not
show the same sensitivity to input frequency. Rather, those children
seemed
to have extracted generalisations from the input. The findings showed
that
children with DS and SLI can present with command of forms like
irregular
past tense appropriate for chronological or mental age, whilst at the
same
time present with a deficit with many grammatical morphemes, including
past
tense -ed and possessive -s. Additionally -and specific to SLI- the
findings support research suggesting that problems with grammatical
morphemes are less obvious in many older children with that condition.

The thesis findings contribute to understanding of grammatical morpheme
acquisition in HI, in DS and in SLI and to knowledge of the underlying
nature of grammatical morpheme acquisition generally. The findings
indicate
that the underlying linguistic system is ultimately regulated by a dual
mechanism and that linguistic abstractions may be more dependent on
specific linguistic experiences than a classic generative account would
predict. Additionally, the research provides direct evidence that data
from
elicitation procedures can be as representative a measure of children's
use
of grammatical morphemes as spontaneous speech data.
--
Peter T. Daniels grammatim@xxxxxxx
.

Quantcast