Please enable JavaScript.
Coggle requires JavaScript to display documents.
"On the different roles of Vs & Cs in Speech Processing & Lg…
"On the different roles of Vs & Cs in Speech Processing & Lg acquisition"
2003 paper by M. Nespor, M. Pena, J. Mehler
1) INTRODUCTION
The notion of result in phonology
(within the tradition of generative grammar)
2 requirements
for a notion to be successful:
1) minimal -
account for all patterns connected with the phenomenon
2) further (advanced)
- be able to explain other phenomena
In particular (a criteria for a result in phonology to be good)
1.1) the result should account for the interpretations of the structures on other layers of lg (morphological, syntactic, lexical)
account for the interpretations of structures generated by generative components (morphology, syntax, lexicon)
1.2) its results should agree with other disciplines with their different methodologies, which also study the linguistic sound systems
The
HYPOTHESIS
of the work
essence
extended:
there is a (partial) division of labour btw Vs & Cs
Vs
main role:
signal (some of)
grammatical properties
related to:
-specific rhythmic class
-syntactic system (universal & subject to parametric variation)
Cs
main role
: distinction btw
lexical
items (lexicon)
short:
Cs & Vs have different functional roles
fuzzy boundaries
the division of labour btw Vs & Cs has some fuzzy boundaries
EX: in particular in the area of
morphology
(there both categories play a role (ex inflectional systems)
2) QUALITY DISTINCTIONS & THE LEXICON
Cs
Thesis:
Cs are especially dedicated to lexical interpretation
is supported by evidence that points to the richness of quality distinctions Cs make \ poorer distinctive power of Vs
EVIDENCE
1) Numerical superiority
of Cs over Vs in most linguistic phonological systems of the world
proof
Proportion of Cs & Vs in lg-es
common
much more Cs \ Vs
Malay (20C:5V)
Italian (24C:7V)
Hausa (32C:5V)
Arabic (29C:3V)
(less common) very little amount of Vs
3V - Arabic, Aranda, Greenlandic
Extremes
(rare) equal \ more Vs
Swedish (16C:17V)
(rare \ exceptions) small number of Cs
Hawaiian (8 CS)
Rotokas (6 Cs)
*Even there there are more Cs than Vs
GENERALLY
Phonological system of most Lg of the world
5 Vs + over 20 Cs
One
reason
for this:
the anatomy of the speech tract
@ the way Cs are pronounced allows more variation
"a larger variety of consonantal than of vocalic segments can be produced by human articulators, such that fairly large changes at times leave the phonetic category unchanged, while other minimal movements may result in a change of category."
Conclusion:
the higher number of Cs in most systems makes them relatively more informative, & their information load may result from this characteristic (but not only from this one)
2)
Cs tend to
disharmonize in a word
, unlike Vs
-in other words:
Cs tend to become more distinctive in one word.
There is a tendency for Cs, belonging to one lexical item (word) to alternate in quality.
Loss of distinctiveness
Differences in the way Vs & Cs lose their distinctiveness
tendency to do that
Usually :check:
How it happens in
2 different phonetic systems
Harmonic Lg-es
1 more item...
Nonharmonic Lg-es
1 more item...
Usually :green_cross:
3) Classical Greek
avoids 3 aspirated Cs in one word
(the Grassmann Law)
2) Arabic
avoids adjacent root Cs produced by the same articulator
1) Japanese
combination of 2 voiced
obstruents
is avoided
When they do...
When
in specific environments
when they undergo weakening
How
1 more item...
Characteristics:
When
usually
(they have a tendency to do that no matter the phonological system of a lg)
How
1 more item...
Cs
Vs
3) Consonant & Vowel clusters
3.1)
Morphological roots
essence
Cs, not Vs, may constitute Morphological roots in some Lg-es
Case of Semitic Lg-es
EX:
Arabic
1 more item...
=> in such lg-es the role of distinguishing lexical meaning rests
exclusively on Cs
3.2)
C & V tiers used in Phonological Representation
the 2 levels of morphological representation
Consonantal tier
motivation:
mainly lexical
Vocalic tier
motivation:
prosody (which signals syntax)
4) Experimental proof
(word recognition & Lg comprehension)
4.1)
Word recognition
Experiment with 'kebra'
Instruction
Pple hear a nonword, and are asked to make it into a real word by changing only 1 phoneme
EX:
kebra
2 more items...
Experiment results
Listeners tend to come up with word 'cobra', rather than 'zebra'
Conclusion
A vowel substitution is easier than a consonantal one
an important detail
on whom
the experiment was conducted
It was carried out with speakers of
Spanish (lg with more Cs than Vs)
Dutch (balanced C-V ration)
Conclusion from that
the more (lexical) distinctive role of Cs is independent of the specific repertoire of a lg \ is not only due to the numeral superiority of Cs in a lg (there is more to it)
4.2)
Lg comprehension
Gedanken experiment
The sentence
without Cs
:red_cross: even with the correct rhythm & intonation, one will be unable to grasp the meaning of the words
Without Vs
:check: able to grasp a few \ if not most lexical items
5)
Cs have
tendency to alternate in quality
, unlike Vs
Proof
1) Words across languages
trisyllabic words with same phoneme in each syllabic nucleus
Vs
easy
to find in many lg-es
Ex
Italian: banana, rotolo (roll)
Turkish: kelebek (butterfly), arkada*s (friend)
Greek: irini (peace), thalasa (sea)
Cs
hard to find
Exception: honomatopeias
Conclusion
Vs
are
not required to alternate in quality
, but do alternate in quantity
Cs
Apart from reduplications & honomatopeias, Cs tend
not to preserve
across one word
Alternate in quality
2) Tongue twisters
Usual essence of TT
combine elements, which are difficult to pronounce > confuse the articulatory program
What they are based on, across Lg-es
Across lg-es TT are based on the similarity &\ro incompatibility of the Cs, comprising a string, not of the Vs
CONCLUSION
=> TT are hard to pronounce, because the Cs \ consonant clusters are too similar to each other > confusing effect (it is unusual & uncomfortable).
A sequence of sentences with similar Vs :red_cross: same confusing effect.
6)
Across Lg-es of
different rhythmic classes
the number of Cs does not vary much as the number of Vs
3 rhythmic lg classes
1) stress-timed
rich in monosyllables
rich syllabic structure
Cs outnumber Vs visibly
Common word = 1 V + 2\more Cs
2) syllable-timed
simpler syllabic structure
Cs outnumber Vs in a less radical way
longer words
Common word = 2\3 Vs + 2-4 Cs
3) mora-timed
very few syllable types
Nu of Cs & Vs tends to be more similar
even longer words
Common word = 2\3 Vs + 2\3 Cs
Conclusion
Cs
tend to be of the
same number
in all the types (2-3 in a word)
Vs
vary in number
depending on the rhythmic class of lg
The fact that across lg-es that belong to dif rhythmic classes, the number of Cs in a word tends to be similar, while that of Vs varies a lot depending on the class >
Cs have less to do with rhythm, than Vs do
limits of the paper:
"the proposal put forward in this paper concerns
exclusively lg-es in which tone does not contrast words
." (it is to be addressed in a separate paper)