Nonparallel recovery in bilingual aphasia: effects of language choice, language proficiency, and treatment.

Abstract

We describe a 57-year-old Russian-Hebrew bilingual aphasic patient who received speech-language therapy in his second language (Hebrew) in the first three-and-a-half months post onset and then in his first language (Russian) for an additional month and a half. He was first diagnosed with Expressive-Receptive aphasia in both languages. After four weeks of treatment in the second language, his language skills improved and he was subsequently diagnosed with Predominantly Receptive aphasia in both languages. Three-and-a-half months post onset, he was diagnosed differently in the two languages: Predominantly Receptive aphasia in Hebrew and Amnestic aphasia in Russian. Following additional six weeks of therapy, this time in his first language (Russian), the patient was diagnosed as Amnestic in both his languages. We present the course of his improvement as seen in four successive evaluation periods in both the treated and nontreated languages, in all language modalities. We address various factors that may have contributed to the nonparallel recovery of the two languages and discuss the relative contribution of spontaneous recovery, therapeutic transfer, language proficiency, language use, and structural relations between the two languages.

Key words

aphasia bilingualism therapeutic transfer treatment

Introduction

Despite early controversy and lingering skepticism, there is now research evidence supporting the assumption that treatment in aphasia facilitates language recovery. Whether treatment in one language contributes to the recovery of a nontreated language in polyglot aphasia, however, remains to be demonstrated. In this paper, we discuss data concerning treatment in polyglot aphasia as we describe the course of nonparallel recovery of the two languages of a Russian-Hebrew bilingual aphasic speaker.

1.1

Polyglot aphasia

Two possible patterns of language deficit and language recovery have been identified in the literature on bilingual and multilingual speakers who had suffered aphasia: parallel and nonparallel (e.g., Albert & Obler, 1978; Charlton, 1964; Paradis, 1977, 1983, 1998). In the more common case, that of parallel deficit and recovery patterns, both (or all) languages demonstrate similar deficits and appear to recover at a similar rate. In some cases of bilingual aphasia, however, one of the languages appears less affected and shows faster recovery; moreover, subtypes of nonparallel recovery have been identified (Paradis, 1977, 1993a). These include successive, or antagonistic recovery, that is, when one language precedes the other(s) (e.g., Halpern, 1941, in Paradis, 1983); selective recovery, in which only one of the languages seems available for recovery (e.g., Berthier, Starkstein, Lylyk, & Leiguarda, 1990); alternate pattern or alternate antagonism, in which bilingual speakers can use one language on one day but only the other language on another day, thus experiencing difficulty with the two languages alternately (e.g., Nilipour & Ashayeri, 1989; Paradis, Goldblum, & Abidi, 1982); and unintentional language switching, in which bilingual speakers are unable to avoid switching between their languages (e.g., Fabro, Skrap, & Aglioti, 2000; Perecman, 1984). Theories have been developed to account for these nonparallel patterns, focusing on two aspects of the phenomenon. The underlying neural and psychological mechanisms that may account for the differential deficits (e.g., Green, 1986, 1998; Obler, 1984; Paradis, 1985; Pitres, 1895) and the variables that determine which language will have an advantage over the other(s) (e.g., Krapf, 1961; Paradis, 1978, 1989; Pitres, 1895; Ribot, 1882).

As early as in 1895, Pitres suggested that nonparallel recovery in bilingual aphasia is not the result of selective destruction of one language, but of its temporary or permanent inhibition. Green (1986, 1998) elaborated on this notion and hypothesized that aphasic bilinguals' ability (or inability) to regulate activation and inhibition processes of the languages is responsible for the differential patterns when they are observed. Green's model of inhibitory control of resources suggests that it is not the languages per se that are affected in cases of nonparallel patterns in bilinguals after brain-injury, but rather the ability to access and inhibit them appropriately; a language must be adequately activated to be selected and adequately inhibited to avoid interference. In a similar approach, Paradis (1985, 1989) proposed the activation threshold hypothesis to account for nonparallel patterns of polyglot aphasia. He hypothesized that each language component needs to reach a certain threshold of activation to be produced. Stimulation and frequent use of a component lower its activation threshold and, in contrast, inactivation raises that threshold. Moreover, frequent activation of a given component (e.g., a word) may reduce activation levels of competing candidates, including components from another language. Aphasia might result in elevated thresholds of a language or of certain language components; inactivation of those components will further lead to high activation-thresholds and production in that language will not be possible. Patterns of selective inactivation and threshold elevation can account for nonparallel patterns of language deficits in polyglots. Alternatively, differential patterns of deficits and recovery can be accounted for by hypothesizing that different neural networks in the brain are associated with the different languages. Indeed, both between-hemisphere (e.g., Gloning & Gloning, 1965) and within-hemisphere (e.g., Ojemann & Whitaker, 1978) differential organization have been proposed. If the two languages are processed by different brain areas, differential patterns such as selective and successive recovery can be explained.

Regardless of whether the differences in the availability among the languages are temporary or permanent, several variables have been identified as potential determiners of the nonparallel pattern. Ribot (1882) proposed that the first language to be learned would be the least affected by brain damage. This may be particularly relevant for bilinguals who learn their languages at different ages (i.e., coordinate bilinguals) as compared to those who acquire both languages simultaneously (i.e., compound bilinguals) (see, e.g., Lambert & Fillenbaum, 1959: Weinreich, 1953). Pitres (1895), in contrast, suggested that the language most used around the time of the onset of the aphasia would be most likely to recover first. Others have suggested additional considerations such as level of proficiency in each language, history of language use (e.g., Abutalebi, Cappa, & Perani, 2001: Albert & Oblel; 1978), and personal preferences (e.g.. Minkowski, 1963).

1.2 Treatment of bilingual aphasia

Beyond these theoretical discussions lie the practical questions: Which language should be chosen as the language of speech-language therapy? What variables interact with treatment of bilingual aphasia? Is there transfer of therapeutic benefits from the treated to the nontreated language? What variables affect transfer? Answers to these questions are critical to the study and application of bilingual aphasia therapy. Researchers have considered different approaches. Krapf (1961) suggested that the preferred language for therapy should be the mother tongue. Others favored the language that appeared to be better recovered (Chlenov, 1948: Hilton, 1980: Lebrun. 1988). Paradis (1983) suggested that the transfer of therapeutic benefits from the treated to the nontreated language is a function of different variables, such as structural distance between the languages, order of acquisition, proficiency and use preonset and postonset, type of aphasia, pattern of recovery, and type of therapy. To date, only a few published studies provide detailed information about selective treatment in bilingual aphasia and possible therapeutic transfer between the two languages. In each of these studies, results may be attributed to effects of spontaneous recovery as well as to those of treatment. We describe four studies below; details are provided in Appendix A.

Fredman (1975) conducted a large survey of 40 coordinate bilingual aphasic patients who had immigrated to Israel as adults and had received speech and language treatment only in their second language (L2), Hebrew. She reported positive effects of the therapy on the patients' communication skills in their first language (L1). Therapeutic benefit was observed in the nontreated languages despite the fact that the majority of Fredman's participants had come to Israel from Eastern European countries and the languages they spoke as L1 were structurally different from Hebrew, the language of therapy. Yet Fredman's results should be viewed with caution because the data were based on questionnaires filled out by the patients and their families, rather than on objective testing. Voinescu, Vish, Sivian, and Maretsis (1977) documented a polyglot who was Greek-born and had mastered Russian, Romanian, and German in addition to Greek. He was reported to have been equally proficient in all four languages (although his preonset reading and writing skills in each language were not determined). Following his aphasia, he experienced language deficits in all four languages, with no substantial qualitative differences in the type of deficits across the languages. He received speech and language therapy in Romanian only (not his mother tongue) and showed improvement in all four languages. The treated language did show greater recovery, but this was only evident in expressive language skills. Voinescu et al. concluded that the therapeutic intervention in one language had beneficial effects on the nontreated languages. Perhaps the most cited study of therapeutic transfer is that of Watamori and Sasanuma (1978). Watamori and Sasanuma documented the recovery of two bilingual English-Japanese patients and reported partial transfer from the treated language to the nontreated language in each case. Both patients were compound bilinguals, one diagnosed with Broca's aphasia, the other with Wernicke's. Both patients demonstrated greater recovery in the language that received direct speech-language therapy (English in both cases). Watamori and Sasanuma pointed out all interesting pattern of parallel improvement in both languages in the two receptive modalities, auditory comprehension (for both patients) and reading comprehension (for the Broca's patient only), in contrast to the greater improvement in the treated language than in the nontreated language evident in the two expressive modalities, writing (for both patients) and oral production (for the Wernicke's only). In addition, Watamori and Sasanuma noted that improvement in writing skills was particularly dependent on direct therapy. The fundamental differences between the two writing systems (English and Japanese) may have contributed to this finding (Paradis. 1993a). However, in another study of an aphasic patient who was proficient in two languages with different orthographies, Khamis, Venkert-Olenik and Gil (1996) reported overall transfer of therapeutic benefits from the treated language, L2 (Hebrew) to the nontreated language, L1 (Arabic). Improvement was seen even in the written modality, despite the different writing systems of Hebrew and Arabic. Khamis et al. hypothesized that, among other factors, the similarity between the morphosyntactic structures of Arabic and Hebrew contributed to the transfer …