Target MC’s response Type of error
2.5.2 Sentences and connected text
M C ’s reading of sentences and connected text is discussed in Druks and Froud (submitted). At first pass, M C’s reading aloud looks almost classically agrammatic, because many function words are omitted (and / or substituted), resulting (amongst other things) in the reading of passive sentences as if they were active. For example:
(9) Targets
a. Sue was kissed by Bill.
Responses
the...Bill.
b. Alex was tickled by his son. Alex was tickling the boy.
However, this classification does not bear a closer examination; as can be seen by the presence of determiners, auxiliaries, and participial endings, M C ’s reading is not characterisable as an omission of all functional morphemes. (More sentence reading data are provided in Appendix VII.)
Rather, it seems as though M C ’s reading of items longer than single words is characterised by his recognition of substantives. He seems to seize on one or two contentful words in the sentence, and then build a phrase or sentence around those. This results in reading performance which is a conglomerate of his reading and spontaneous speech. Consider the following examples:
(10) Targets Responses
a. The horse is carrying the cat. The horse carries the cat. b. I wonder when spring will come. I wish spring would come.
c. John came to the party with his John likes the party - can't do these friend Peter. things - and a friend.
d. Mary phoned John but he was out. John phoned Mary but no answer. e. Peter ate all the jam tarts. Peter likes jam tarts.
It is noticeable that function words which MC is unable to read when they are presented to him in isolation are very often appropriately used in reading of larger syntactic units; this is also a characteristic of his spontaneous speech, and can essentially be explained in the same way. Because M C’s access to lexical representations of substantives is assumed to be intact (as witness his very good performance on lexical decision tasks - section 2.7) it is possible that he has access to the representation of the extended projection of substantive categories. W hilst this is not as much information about the syntactic representation of function words as is available to the normal computational component, it is more than is available to M C’s^ computational component when it is required to deal with function words in isolation. In a sense, M C ’s faulty morphological realisations of functional categories are supported by the presence
of information about the extended projection of substantives. If MC can recognise a couple of substantives from the sentence he is reading, he can also boost his deficient representation of the associated functional categories, and in some cases this is sufficient information for him to produce a fully grammatical sentence - though the extent to which this resembles the target is extremely variable. M C ’s reading of connected text (from Neale 1989) can be seen to support this conjecture: >
(11)
Target
The lions’ final act was in progress. Jack stood waiting to clear the ring. The thunder outside the circus tent had made the lions restless. Suddenly Tina, the lion trainer, stumbled. Her whip fell. The youngest lion sprang towards her. Jack leaped swiftly inside the cage, cracking the whip with great skill. His prompt action enabled Tina to regain control quickly. After that brief adventure. Jack decided on his future work.
Attempt
The lions in progress in the ring. Jack...in the door, no, it is waiting in the...erm ...rin g . The thunder is outside, the lions are restless. Tina falls and the trainer stumbled. The whip fell. The lion...the, the...erm , erm', the something lion prang (=sprang?) towards me. Er, Jack swiftly came in the cage with the, erm ....w hip with control. His br..His cage otion tembled to, to control quickly. After the venture Jack, Jack decided we his work were done.
It is apparent that M C’s reading of connected text contains some neologisms and articulation errors. These are rare in his spontaneous speech and in his reading of single words and sentences, and it is therefore assumed that they result from increased pressure on his reading performance. MC is very aware of his errors and is sometimes a little embarrassed by them, which can increase his frustration and lead to the production of more errors; reading of connected text certainly yields more such difficulties than reading o f shorter stimuli.
It is clear, however, that M C ’s poor reading of function words is considerably improved by the presence of a syntactic frame, presumably due to the availability of
syntactic information from the extended projection of the substantives present in a sentence. In order to establish the validity of this conjecture, MC was given some specially designed tasks (reported in Druks and Froud submitted). The first contained sentences with a function word in roughly medial position. The function word was emphasised (by being printed in a larger, bold font), and MC was asked to look at the whole sentence but only read the emphasised word. For example:
(12) a. I wonder
why
the sky is blue.b. Peter decided
that
he would stay at home. ' c. The burglar was caughtby
the policemen.MC had great difficulty in responding to this task, and needed many reminders and practice items. Nevertheless, he produced 14 out of 40 function words which were appropriate to the sentence; however, he still did not seem to be actually reading them. For instance, he responded to John went to work AFTER eating breakfast with ‘before’ - which was scored as appropriate because it fits into the syntactic frame, but is nevertheless incorrect. On the strict criterion, 12 of M C ’s responses actually matched the target function word - which is still much better than his overall reading of function words in isolation - 30% correct as compared to 11.67% (which is the percentage of correct responses reported in Appendix V). It seems clear from these results that M C ’s production of function words is supported by the presence of a syntactic frame. (The results o f this task are given in Appendix VUI.)
M C ’s reading of longer items than single words has not been investigated further for the purposes of this study, because it results from a complex interaction of linguistic and extra-linguistic factors which it is very difficult to identify. In addition, having established that the presence of a sentence frame is likely to support M C ’s reading of function words which he cannot read in isolation, the only way to really investigate the processes and deficits which may underpin M C’s difficulties with function words is to ensure, as far as possible, that he is not making use of support mechanisms, linguistic or otherwise. For these reasons, single word reading forms the major focus of the present study.