Summary of Harwintha's PhD dissertation

Agrammatic aphasia is a complex of language problems that occurs after damage of the language areas of the brain’s left hemisphere. The areas usually are or include Broca’s area (Brodmann Areas 44 and 45). Generally, agrammatic speakers have problems with grammatical features of language in production and comprehension. In experimental production and spontaneous speech studies of agrammatic speakers of various inflectional languages, such as Dutch, English, German, Greek, and Turkish, time reference has been found to be a weak spot. However, more data are needed to see if the time reference problems are due to difficulties in inflecting the verbs, in referring to a certain time frame (e.g., past), or they are due to some other reasons not yet known.

It is interesting to investigate time reference in Standard Indonesian (SI) for three main reasons. First, time reference in SI is done in two ways, both of which are words or free standing morphemes, not bound like inflectional morphemes. The first way is through the syntactic aspectual adverbs (e.g., sudah, sedang, and akan) and the second by lexical adverbs of time (e.g., baru saja, sekarang). By investigating them, we tease apart the confound of time reference and inflection. Second, related to the first reason, by studying SI we can see if there is a central problem faced by agrammatic speakers when they need to refer to time. In other words, we want to know if the problem arises regardless of how time reference is done linguistically. The third, less related, but not less important reason, is the fact that SI has hardly been studied. Belonging to a different language family than other languages reported so far in aphasiological, neurolinguistic, or psycholinguistic literature and having a large number of speakers make SI a language that can contribute considerably to the field(s).

The work discussed here is a part of the cross-linguistic study on time reference in the Neurolinguistics group in Groningen. Apart from looking to time reference in SI, we also evaluated the test battery which is also used in more than fifteen languages. Thus, our results also informed this large scale study as to whether the test is valid to be used for investigating time reference in SI.

Chapter 1 gives background information about SI that is needed to understand the following chapters. Some theoretical accounts related to time reference in agrammatic aphasia were presented, two of which were chosen to be tested in the current project. Four general research questions were formulated at the end of this chapter.

Will the Standard Indonesian (SI) aspectual adverbs be difficult for the agrammatic speakers?

The weak syntax model predicts that they should be difficult because they are discourse-linked. Meanwhile, the PADILIH predicts that only the perfektif aspectual adverbs will be difficult.

Will the lexical adverbs be difficult for the agrammatic speakers? 

Classically, these adverbs have never been documented as impaired in the literature. However, the fact that they are used to refer to time and are discourse-linked predicts that they can be difficult for the SI agrammatic speakers.

Are the difficulties caused by these two kinds of adverbs similar or different? 

If they are different, are the differences related to the modality (spontaneous speech, comprehension, and production)?

 

Chapter 2 aims to characterize agrammatic speech in SI because there was no test or method to screen SI-speaking participants with agrammatic aphasia who could participate in our time reference studies. Based on some variables that have been widely published in the literature and some variables that are unique to Standard Indonesian, we concluded that SI agrammatic speech is characterized by short and syntactically simple sentences, slow speech rate, low proportion of particles in general and syntactic particles in particular, problems in realizing grammatical objects after accusative markers, possible problems with derivational affixes and reduplication, and no problems with verbal predicates and derived word order (passives). The participants of this study, who had agrammatic speech, participated in the time reference studies discussed in chapters 3, 4, and 5. 

In chapter 3, the study investigating the relationship between the diversity of verbal predicates (as measured by Type Token Ratio) and the aspectual adverbs produced with the verbal predicates is presented. When the performance of the agrammatic participants was compared among one another, the results showed a trade-off between the two variables. On the one hand, the agrammatic speakers who had higher Type Token Ratios produced aspectual adverbs less frequently.  On the other hand, the agrammatic participants who had lower Type Token Ratio produced aspectual adverbs more frequently. The agrammatic speakers who produced aspectual adverbs less frequently did not overuse lexical adverbs to compensate. Therefore, research question number 1 can be answered affirmatively (the SI aspectual adverbs are difficult) and there is a trade-off between the diversity of verbal predicates (as measured by Type Token Ratio) and the percentage of aspectual adverbs produced with the verbal predicates, without an overproduction of lexical adverbs of time to compensate the need to refer to time.

The dissertation continues to chapter 4 which presents the comprehension study. The non-brain-damaged (NBD) participants performed at ceiling in the TART comprehension test. The agrammatic participants performed significantly worse than the NBDs at the group level in both the aspectual adverbs and lexical adverbs of time tasks. At the individual level, five agrammatic participants scored significantly worse than the NBDs in both tasks. For the scores of the agrammatic participants, there was a strong correlation between the scores on both tasks. In other words, they were similarly impaired in understanding sentences containing aspectual adverbs and sentences containing lexical adverbs of time. At both levels and both tasks, no time frame was selectively impaired. We attributed this to the fact that in SI the use of the adverbs is optional and these adverbs are used when context does not provide enough time reference information. Therefore, all the adverbs are equally vulnerable in comprehension. Based on these results, we can answer research questions 1 and 2. In comprehension, the time reference problems affect both ways of referring to time and there was no statistically attested difference posed by the aspectual adverbs and lexical adverbs of time. 

The production study is presented in chapter 5. With this study, we wanted to answer all four questions, especially regarding the comparison with the results of the comprehension study. Here, the results showed that the agrammatic participants also had problems producing sentences containing aspectual adverbs and lexical adverbs of time. In production, there seemed to be a qualitative difference between the two kinds of adverbs that made the lexical adverbs more susceptible to be substituted by aspectual adverbs than the other way around. We suggested that this was because of the referentiality of the lexical adverbs of time which made the agrammatic participants integrate between linguistic level and context. The non-referentiality of the aspectual adverbs did not necessitate this integration. However, since the NBDs did not perform at ceiling, suggesting that the production test is not the optimal means for SI, we did not draw further conclusions.

Chapter 6 presents a comparison across the four studies discussed in the previous chapters. Five agrammatic participants participated in the three time reference studies and three of them were impaired in all those studies. This means that these five agrammatic participants showed the trade-off in chapter 3, but only 3 were impaired in comprehension and production. 

In chapter 7 the results were discussed in relation to the literature and some suggestions for adaptation of the two theories were suggested. 

There is a selective difficulty in referring to the past when time reference is obligatorily marked (e.g., in English, Dutch).

2. Referring to the past is not selectively difficult when time reference is optionally realized and is used to link the event to discourse.

3. In the case of optional time reference markers, the markers that need more integration with discourse level representations will be more impaired than the markers that require less of this, at least in production. 

This revision needs to be tested in other languages. In the interest to advance knowledge in aphasiology and to inform aphasia therapy, we believe our results encourage more research to be carried out in Standard Indonesian, other dialects of Indonesian, and other Austronesian languages.