summary on plural

[Brian MacWhinney]

Dear Info-CHILDES,
     I received 22 responses to my posting regarding evidence for  
early comprehension of the plural.  I have compiled these responses  
into a set of pointers to the current and forthcoming literature  
which I am attaching here.  I have only been able to read about three  
of the papers mentioned here so far, but I have already learned quite  
a bit.  First, it  is clear that the plural is not a uniform  
category, as Roeper, Pérez-Leroux and other argue.  Second, although  
the initial application of preferential looking to this topic was  
unsuccessful, there appear to be some recent successes.   In  
particular, studies that demonstrate productive control of the plural  
with novel nouns in comprehension provide us with an important  
landmark in terms of understanding the overall course of learning of  
morphology.  Exactly where to place that landmark vis a vis  
production is perhaps not yet entirely clear. Third, it would be  
helpful to compare the learning of the English plural with more  
transparent plural markings such as Hungarian -ok or Turkish -ler.

--Brian MacWhinney

1.  Dave Barner, Harvard

Several studies at Harvard have investigated this.

Kouider, S., Halberda, J., Wood, J. & Carey, S.  Acquisition of  
English number marking: The singular-plural distinction.  Language  
Learning and Development, 2, 1-25.   Using preferential looking, this  
study found that 24-month olds look reliably at the plurals of novel  
words if the noun and the verb are plural, but not if only the noun  
is plural.

A second study, that I performed (Barner, Thalwitz, Wood, Yang &  
Carey, under review), shows that (1) children successfully search  
longer for 4 objects than for 1 at 22-months but not earlier  
(following up on Lisa Feigenson's finding that younger children can  
track only up to 3 objects); and (2) success at this 1 vs. 4 task is  
driven by children whose parents report they use plural nouns in the  
MCDI.

Finally, a third study (Wood, Kouider, Halberda & Carey, under  
review) uses the same box reach task. The experimenter looks into the  
box and claims that either "there is a blicket" or "there are some  
blickets". Again, 20-month olds succeed and 24-month-olds fail.

So, the sum of these results correspond with the MCDI production  
norms: children seem to be acquiring the link between s-p morphology  
and the underlying conceptual distinction sometime between 20- and 24- 
months, and in one case at 22-months exactly. MCDI norms indicate  
that 50% of children produce the plural at 22-months.

Subsequent work has investigated whether the underlying conceptual  
distinction might be available to children prior to this (Barner,  
Kibbe, Wood & Carey, in prep). We've also looked to see whether a  
privileged distinction might exist between a "single individual" and  
"more than one" in rhesus monkeys (Barner, Wood, Hauser & Carey, in  
prep). The last two studies have played with spatio-temporal cues  
like common motion to elicit the representation of sets without  
invoking numerical representations. They may be totally unrelated to  
singular-plural in language - it remains to be seen what the link is.

Another line has been to investigate French kids, to see if cross- 
linguistic differences affect age of acquisition (in progress by  
Kouider, Feigenson and Halberda) and to look at the conceptual  
distinction in Mandarin (in progress with Peggy Li, Shu-Ju Yang and  
Susan Carey) and Japanese children  (in progress by me, Tamiko Ogura,  
Barbara Sarnecka, & Susan Carey). In a year or so we should know how  
all of these pieces fit together!

2.  Kamil Ud Deen, University of Hawaii at Manoa

While a grad student at UCLA in Nina Hyams' lab, we tried something  
like what you are suggesting, with no result.  The reason was that  
the pictorial stimuli, while seemingly very transparent and amenable  
to the methodology, biased children towards plural responses.  So for  
example, when presented with a banana on one side and a bunch of  
bananas on the other side, because many bananas are inherently more  
interesting, we got a skew towards plural responses.  This was true  
for inanimate as well as animate pictures, despite various attempts  
to address the problem.  I'm not trying to point out how dumb and  
unimaginative we were (and I'm sure someone smarter than us has  
managed to get around this), I just wanted to point out that is is  
not the case that this methodology lends itself so well to the  
testing of plurality.

***  MacWhinney comment:  Erik Thiessen (CMU) told me that Anne  
Fernald ran into similar methodological problems with this about 10  
year back, supporting what Kamil says.  Of course, this new report  
from Kouider et al. suggests that this can be overcome.

3.  Lise Menn, Colorado

For a single-child data point, you can look at:
Peters, A., and Menn, L. (1993) False starts and filler syllables:  
Ways to learn grammatical morphemes. Language  69:4 (1993). pp. 742-777.
Daniel was still demonstrably not comprehending the plural marker at  
2;2.19, but started to produce it (and presumably to comprehend it??)  
about two weeks later.

**** MacWhinney comment:  I would say that this is the closest we  
have so far to an estimate of the comprehension-production lag for  
this morpheme.  This estimate seems to fit with similar single-case  
anecdotal data from Brown.  Obviously we would need to have this  
confirmed experimentally, but it may well be true that the gap is not  
as huge as one might expect.  The discussion of the semantic problems  
with the plural from Roeper and Pérez-Leroux would further support  
the idea that the plural doesn’t just emerge full-fledged in either  
comprehension or production.  Still, emerge it does.

4.  Recent unpublished studies by Soderstrom, Newman, and Plunkett.   
Work currently under review.

Kouider et al. mention two recent unpublished studies that examined  
this issue using preferential looking.  These are from Melanie  
Soderstrom and Rochelle Newman, both of who also replied directly to  
me.  Rochelle’s study was presented at SRCD

Schnoor, B. C., & Newman, R. S. (2001). Infants' developing  
comprehension of plurals, 2001 biennial meeting of the Society for  
Research in Child Development.

Melanie’s study is her unpublished doctoral dissertation from Johns  
Hopkins in 2002.

Melanie’s note to me suggests that there is “sensitivity to -s  
inflection (both plural and 3rd singular) in 16-month-olds.”

Kim Plunkett (Oxford) found evidence for early sensitivity using  
novel words.  Kim included a copy of his unpublished article on this,  
but I accidentally failed to pull out the attachment.  My apologies  
to Kim.


5.  Michael Tomasello, MPI Leipzig

The youngest children to add English plural -s in a production  
experiment with novel verbs are, to my knowledge, 21-22 months (4 of  
10 children at least once).

Tomasello, M., Akhtar, N., & Dodson, K., Rekau, L. (1997).  
Differential productivity in young children's use of nouns and  
verbs.  Journal of Child Language, 24, 373-87.

I know of no preferential looking experiments examining this with  
novel verbs.


6.  Dick Weist: Fredonia

As I understand your message, you are ultimately interested in the  
comprehension - production lag, and you are approaching the problem  
through the singular - plural distinction because of linguistic and  
methodological transparency. I might suggest that you can also  
approach this problem by investigating temporal and spatial reference  
with a similar methodological advantage.  Here are three references  
with one on the comprehension side of the issue and two on the  
production side.  I have added a forth reference that places this  
work in an integrated and larger context.

Weist, R.M., Atanassova, M., Wysocka, H., & Pawlak, A. (1999).   
Spatial and temporal systems in child language and thought: A cross- 
linguistic study.  First Language, 19, 267-312.
Weist, R. M., Pawlak, A., & Carapella, J. (2004).  Syntactic-semantic  
interface in the acquisition of verb morphology.  Journal of Child  
Language, 31, 31 – 60.
Internicola, R. & Weist, R. M. (2003).  The acquisition of simple and  
complex spatial locatives in English: A longitudinal investigation.   
First Language, 23, 239 -248.

Weist, R. M. (2002).  Space and time in first and second language  
acquisition: A tribute to Henning Wode (pp. 79–108).  In P.  
Burmeister, T. Piske, A. Rohde (Eds.) An integrated view of language  
development: Papers in honor of Henning Wode. Trier:  
Wissenschaftlicher Verlag Trier (WVT).

7.  Javier Aguado-Orea: Nottingham

Casla, M, Aguado-Orea, J & Pine, J. (2005). Eliciting frequent and  
infrequent verb forms in Spanish: An experimental study of the  
acquisition of inflectional morphology in Spanish. Paper presented at  
X International Congress for the Study of Child Language. Berlin,  
July 2005.

The idea was to ask 3 year old children "qué hace" and "qué hacen"  
before a series of images displaying different characters performing  
a number of familiar actions. We wanted to know if the proportion of   
plural agreeing inflections (e.g. juegan) as RESPONSES to a plural  
agreeing question (i.e. hacen) was smaller than the proportion of   
singular agreeing inflections (e.g. juega) as responses to a singular  
agreeing question (e.g. hace). The effect was found, and it was  
significantly smaller than in adults (used as controls).

This result seems to be consistent with previous results indicating  
higher error rates and lower productivity for the USE of plural  
agreeing verb inflections in Romance languages.
Two biased examples follow:
Aguado-Orea, J. (2004). The acquisition of morpho-syntax in Spanish:  
Implications for current theories of development. Unpublished PhD  
Thesis, The University of Nottingham, Nottingham.
Rubino, R. & Pine, J. M. (1998). Subject-verb agreement in Brazilian  
Portuguese: what overall error-rates hide. Journal of Child Language,  
25, 35-59.

Note that I have capitalised 'RESPONSES' and 'USE'. I know that you  
wanted something about *comprehension*, but, the fact of the truth is  
that Spanish learning children use AT LEAST ONE plural agreeing  
inflection from very early, so there must be something going on  
there, since they have problems to be fully productive and accurate  
later on...

8.  Tom Roeper

Plurals are one of the most interesting and inherently difficult  
things that a child must master.  It is not clear when its properties  
are fully grasped at all.  To consider what the child faces, parents  
easily say to a child, holding up a single banana,
    "do you like bananas?"
with a generic reference in mind, but how does the child know that?
An interesting paper on this is at the UMass website by Sauerland,  
Anderson, and Yatsushiro.   Following work by Anne  Vainnikka, they  
asked children questions  like:
             Does a dog have tails?
Try it! Six year olds regularly say "yes".

In my forthcoming book "The Prism of Grammar" (MIT) I devote a long  
chapter to the topic.

MacWhinney comment:  Ana Teresa Pérez-Leroux makes very similar  
points in this article:

Pérez-Leroux, Ana T.  (2005) Number problems in children. Proceedings  
of the 2005 annual conference of the Canadian Linguistic Association.


9.   Some additional notes and pointers:

Gelman, S. A., & Raman, L. (2003).  Preschool children use linguistic
form class and pragmatic cues to interpret generics.  Child Development,
24, 308-325.

Chapter 8 in:  Gelman, S. A. (2003).  The essential child:  Origins of
essentialism in everyday thought.  New York:  Oxford University Press.

Deacon, S. H., Lalji-Samji, N., Leung, D., & Werker, J. F. (2004, May  
9). 'More or less': The specificity of 18- and 24-month old infants'  
knowledge of the plural. Paper presented at the XIV Biennial  
International Conference on Infant Studies, Chicago, Il.

Twila Tardif (Michigan) provided this  further comment on detection  
of the plural in the Werker frameword:

Amber was in a Janet Werker study when she was about 18 months and  
had trouble with "keet" vs. "keets" but not with "keetsu" (Japanese  
version of plural).  She knew no Japanese, but had a huge vocabulary  
for an 18-month-old in both English and Mandarin at that point so I  
was shocked that she didn't make the discrimination.  This fits with  
Janet and her colleague's ideas, though, about the salience of the "s”.

Helen Deacon at Dalhousie may be working on this.
Alison Jolly at Oxford may be working on this.
K. Miller & C. Schmitt in Language Acquisition
Recent work by Jill Devilliers.
Szagun, G. (2001). Learning different regularities:  The acquisition  
of noun plurals by German-speaking children. First Language, 21(62,  
pt. 2), 109-141.