Tutorial

Bilingual Aphasia: A Theoretical and Clinical Review

Bonnie Lorenzen Laura L. Murray
Indiana University, Bloomington

Purpose: To provide an overview of the potential bilingual client population in the United States, present current neurolinguistic and psycholinguistic views of bilingualism in adults with and without aphasia, review related bilingual aphasia recovery patterns and the factors that might influence these recovery patterns, and provide insight into diagnostic and therapy procedures for addressing the needs of bilingual clients with aphasia. Method: A review of the literature was conducted to summarize and synthesize previously published research in the area of bilingual aphasia, highlight unique aspects of aphasia recovery, assessment, and treatment, and identify areas in need of future research.

Conclusions: Despite a growing understanding of bilingualism and the various recovery patterns identified with bilingual aphasia, there remains a dire need for empirically validated management techniques, particularly in terms of determining which language to target, identifying which aspects of various languages are most vulnerable to insult as well as most responsive to treatment, and establishing how to exploit language similarities to maximize treatment efficiency.

Key Words: aphasia, bilingual, recovery, assessment, treatment

n the United States (U.S. Bureau of the Census, 2003), there is a growing prevalence of bilingual speakers, that is, individuals who use two or more languages in their everyday activities (Grosjean, 1989). Speech-language pathologists (SLPs) can therefore expect referrals for an increasing number of bilingual clients who demonstrate a variety of acquired cognitive and communicative disorders, including aphasia. For instance, approximately 45,000 new bilingual aphasia cases are expected per annum in the United States (Paradis, 2001). With this inevitable increase in our bilingual caseloads comes the need to determine effective and efficient assessment and treatment protocols that take into consideration the unique needs and skills of bilingual versus monolingual clients. Accordingly, to increase awareness and understanding of the theoretical and applied issues surrounding service provision to bilingual clients, this article reviews the existing research by discussing (a) the potential bilingual client population in the United States, (b) definitions of bilingualism, (c) neurolinguistic and psycholinguistic views of bilingualism in adults with and without aphasia, (d) bilingual aphasia recovery patterns and the factors that might influence these recovery patterns, and (e) diagnostic and therapy procedures for addressing the cognitive and communicative needs of bilingual clients with aphasia.

The Bilingual Client Population

Demographics
Census data indicate that individuals who speak more than one language represent one of the fastest growing segments of the U.S. population. For example, on the 2000 U.S. census, 47 million individuals 5 years of age and older spoke a language other than English at home (most frequently Spanish followed by Chinese and French), representing a 14% increase since 1990. Of the 28 million who speak Spanish as their primary language, roughly 50% and 90% reported speaking English very well or with no difficulty, respectively, making the majority of this population bilingual. Along with this growth, there are increasing numbers of elderly individuals. For example, the elderly Hispanic population is the fastest growing subgroup in the United States (American Speech-Language-Hearing Association [ASHA], 1991). Rising numbers of elderly individuals are important because of the increased risks of acquiring cognitive-communicative disorders, including aphasia, associated with aging. Whereas to many SLPs it may seem remote to have bilingual clients on their caseload, Laganaro and Overton Venet (2001) pointed out that bilingual speakers are already common in large urban outpatient settings, and all U.S. regions have experienced dramatic increases in the population who 299

American Journal of Speech-Language Pathology Vol. 17 299317 August 2008 A American Speech-Language-Hearing Association 1058-0360/08/1703-0299

speak a language other than English, with the Midwest only slightly behind the South and West in change. Furthermore, these individuals reside in not just large urban areas but also less densely populated regions. For instance, Indiana, between the 1990 and 2000 censuses, experienced a 47% change in Spoke a language other than English at home, rising from 245,826 to 362,082 individuals, with the most significant increases in rural rather than urban areas (Indiana Commission on Hispanic/Latino Affairs, 2005).

understanding the nature of bilingualism and the theoretical models forwarded to explain bilingualism.

Defining Bilingualism
Debate persists on how best to quantify and qualify, and thus define, bilingualism (Baker, 1993; Grosjean, 1989; Mcnamara, 1969; Paradis & Libben, 1987). Of the various definitions, Grosjean (1989) characterized bilingual speakers in perhaps the most realistic terms as those who speak two or more languages in daily life. Perfect knowledge of both languages is not required; instead, people use different languages for different purposes or life domains and consequently have different levels of proficiency within their languages across those domains. For example, in the United States, many immigrants are communicatively competent in their second language (L2), English, for work purposes, but may have quite limited English proficiency in other life domains (e.g., church, home). The terms early and late are often used to describe bilingual speakers language competency, with early bilingual speakers (i.e., those who acquire their languages prior to adolescence) being assumed to have greater proficiency in L2 than late bilingual speakers (i.e., those who acquire L2 after adolescence; Ardila, 1998). Whereas age, sequence, and method of acquisition offer some suggestion of proficiency, there remains great variation, even among speakers with similar acquisition histories. Other factors such as the language used in the school system, personal and social attitudes (e.g., cultural identification and loyalty), and family can affect use and maintenance of both the earlier language (L1) and the later learned L2, even if L1 is used at home (Ardila, 1998; Muoz & Marquardt, 2003). Furthermore, cultural identity can vary geographically and across generations and socioeconomic levels, adding to variability. In addition to varying uses of their respective languages, bilingual speakers may engage in a behavior known as code switching: alternating uses of ones languages to add communicative intent, emphasis, or emotional value (Muoz, Marquardt, & Copeland, 1999). Appropriate code switching requires linguistic competence in identifying and using languages appropriate to communicative interactions. Codeswitching constraints include environmental, social, and personal factors, grammatical principles, and level of education. The acceptability of code switching also varies across communities: Frequent code switching is considered appropriate in some communities but uncommon and less accepted in others. For example, Fabbro (2001a) observed nominal code switching in Italian-Fruilian speakers. Additionally, Muoz et al. (1999) identified and described a number of switching patterns and found that both speakers with and without aphasia exhibited patterns of code switching typified as disordered; these data suggest substantial variability in what is normal across local speech communities. Accordingly, for various reasons, bilingual speakers can differ greatly in their respective uses of and proficiency in their languages. To understand further bilingualism, models used to explain the neurological, linguistic, and cognitive underpinnings of bilingualism are next reviewed.

Health Factors
SLPs working with the aphasia population need to be aware of the risk factors for conditions that cause aphasia not only to provide appropriate client and family counseling and education but also, ideally, to educate the general public and facilitate prevention of aphasia and other neurogenic communication disorders. As Latinos are currently the largest minority in the United States (American Heart Association [AHA], 2005), health concerns particular to this community will be reviewed to exemplify the pressing need for establishing efficacious aphasia assessment and treatment protocols for bilingual clientele. Risk factors for stroke, the primary cause of aphasia, include diabetes, smoking, obesity, high cholesterol, and sedentary lifestyle (AHA, 2005). Whereas current statistics indicate that Latino individuals age 18 years and older have slightly lower rates of heart disease, hypertension, and history of stroke compared with their non-Latino White peers, they are twice as likely to develop an ischemic stroke due to inactivity, obesity, and diabetes. Likewise, among Mexican Americans, the fastest growing Latino group in the United States, stroke incidence is slightly higher at 1.63% compared to 1.36% for their non-Latino White peers (Morgenstern et al., 2004); transient ischemic attacks at younger ages also are more frequent in this group of Latino individuals. Compounding this increased stroke risk in Latino adults is reduced access to appropriate health care services. In a study controlling for age, gender, income, education, insurance, and residence location, Latino individuals were 33% less likely to receive all necessary health care services compared with non-Hispanic White individuals (U.S. Department of Health and Human Services, 2004). Several factors may contribute to poor health care access, including a lack of health insurance, interpreters, and translators (Indiana Commission on Hispanic/Latino Affairs, 2005). The implications for SLPs working with this population are many. They need to be prepared for clients who may not have received expected levels of care acutely or postacutely, or who may not have access to the full range of typical stroke care services. Significantly, SLPs may find themselves needing to advocate for client care, placing more effort into education, and taking on a larger than usual case management role. In summary, the above statistics highlight the rapidly growing bilingual population in the United States and further underscore that many of these individuals are at risk for developing medical conditions that can produce aphasia or other cognitive-communication disorders. Consequently, there is a growing need to ensure that SLPs are adequately prepared to serve the bilingual population. This includes

300 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

Models of Bilingualism
Neurolinguistic Aspects of Bilingualism
Within the field of neurolinguistics, language is viewed as a modular construct in which each linguistic function depends on the involvement of several subcomponents (e.g., spoken output involves concept retrieval, lexical selection, syntactic organization, and oral-motor programming), and each of these subcomponents can be localized in different parts of the brain (see Fabbro, 1995, for a comprehensive review). With respect to bilingualism, neurolinguistic research has focused on determining where different languages are localized and whether all of a speakers languages are localized in similar or distinct areas of the brain. Thus far, separate, shared, and amalgamated representation theories have been put forth to explain the neural organization of bilingual speakers languages. Whereas initially researchers hypothesized that languages resided separately in the brain, this view was replaced with one in which languages shared representation (Fabbro, 2001b). In the early, shared representation approach, all languages were hypothesized to rely upon the same cerebral areas (Fabbro, 1995); if this was the case, however, all languages should be equally damaged following an insult. Because a substantial empirical literature contradicts this predicted outcome and highlights heterogeneity in the recovery patterns of bilingual individuals with aphasia (e.g., Fabbro & Paradis, 1995; Paradis, Goldblum, & Abidi, 1982), researchers have proposed additional factors (e.g., quantity of input from the respective languages following injury, emotional experiences in the languages, language use after injury) that might contribute to differential L1/ L2 recovery (Aglioti, Beltramello, Girardi, & Fabbro, 1996; Goral, Levy, Obler, & Cohen, 2006; Green, 2005; Minkowski, 1927/1983). Combining shared and separate representation views, the amalgamated hypothesis asserts that languages share some areas of the brain but also retain some separate neural areas. Although there are mixed findings in terms of both the cerebral (i.e., which hemisphere supports the language or languages) and intrahemispheric neural representations of language in bilingual speakers (i.e., which structures within a given hemisphere support the language or languages; Paradis, 2004), most research supports the amalgamated view. To illustrate, in most neuroimaging investigations, no laterality differences have been found between monolingual speakers and bilingual speakers (Hernandez, Dapretto, Mazziotta, & Bookheimer, 2001; Hernandez, Martinez, & Kohnert, 2000; Kim, Relkin, Lee, & Hirsch, 1997). Right hemisphere damage in bilingual speakers, consequently, will result in the same patterns of cognitive-communication deficits that monolingual speakers experience, and bilingual speakers who suffer left hemisphere damage are at risk for aphasia onset (Paradis, 2004). Shared lateralization, however, does not imply an absolute overlap of languages or sharing of the exact same neuronal circuits. Whereas research suggests overlap of L1 and L2 (Hernandez et al., 2000, 2001; Kim et al., 1997; Perani et al., 1998), investigators continue to explore factors that influence the extent of this overlap. Age of language acquisition is one proposed factor. For example, functional magnetic resonance

imaging ( f MRI) studies of early Spanish-English bilingual speakers completing naming tasks found no activation differences between L1 and L2, indicating significant neural overlap (Hernandez et al., 2000, 2001). In contrast, Kim et al. (1997) found similar L1-L2 activation in some areas but different activation levels in others for late L2 learners engaged in silent sentence generation tasks, suggesting that age of acquisition plays a role in the pattern of intrahemispheric cerebral activation. Proficiency level, however, may have interacted with age of acquisition in the Kim et al. (1997) study and, in fact, is another proposed factor. For example, using listening comprehension tasks, Perani et al. (1998) found no differences in the activation patterns of early and late bilingual speakers when both groups demonstrated high proficiency in both languages. De Bleser et al. (2003) used positron emission tomography (PET) to examine activation patterns during a picture-naming task. The participants were native Flemish/Dutch speakers who had good proficiency in their L2, French. Although De Bleser et al. found minimal activation differences among naming of cognates in L2, cognates in L1, or L1 noncognates, they did observe a different activation pattern (i.e., left inferior frontal and temporo-parietal areas) when participants named L2 noncognates. They concluded that recruitment of additional neural areas was necessary when producing words in the less proficient language. Accordingly, one theory of second language acquisition, the convergence hypothesis, states that regardless of age of acquisition, as proficiency increases, so too will the level of neuroanatomical convergence (Green, 2005). Language learning modality may also influence the neural mapping of bilingual speakers languages by evoking participation of different memory stores (Paradis, 2004; Ullman, 2001). Implicit memory, more commonly referred to as procedural memory, relates to automatic processes completed within nominal awareness, whereas explicit memory, also known as declarative memory, involves controlled processes carried out at the conscious level (Ullman, 2001). These two memory systems employ distinct cerebral regions: Procedural memory is primarily associated with left frontal and basal ganglia structures, whereas declarative memory is primarily associated with bilateral temporal lobes structures. Both types of memory have been implicated in language learning and use. Implicit memory appears key to L1 development, which occurs during childhood; explicit memory has been implicated in factual learning and the formal language learning that typically occurs during L2 acquisition. Automatic processing, however, is not limited to L1 or languages acquired during childhood. That is, with extensive L2 practice, speakers may develop implicit competence in which L2 can exploit procedural memory and speakers can use the L2 automatically, without explicitly recalling the rules that support accurate comprehension or production (Segalowitz, Segalowitz, & Wood, 1998). Nevertheless, languages learned through formal study (typically L2) may have wider cerebral representation and utilize declarative memory more than L1; in contrast, learning and using L1 more often exploits procedural memory and thus may have a more focal subcortical representation. As these two memory systems employ different cerebral pathways, one or the other may be selectively affected by

Lorenzen & Murray: Bilingual Aphasia Review

301

pathology. With respect to aphasia recovery, if metalinguistic knowledge is available but implicit linguistic competence has been compromised, as in the case of subcortical damage, a paradoxical recovery pattern may occur: better recovery of a weaker L2 due to compensatory reliance on more extensive, explicit knowledge of the L2 (Paradis, 2004). Indeed, Aglioti et al. (1996) described a bilingual client who subsequent to a subcortical stroke demonstrated severe L1 (i.e., Venetian) production deficits despite good L2 (i.e., Italian) production skills and similar degrees of L1 and L2 comprehension involvement. One explanation of this recovery pattern is that the subcortical lesion location caused greater damage to the L1, leaving the explicit memory system more intact and available for L2 processing. It should be noted that this is only one possible explanation, as Green (2005) suggested that the basal ganglias role in controlling competition and pathological switching (vs. implicit memory) could also produce this recovery pattern. That is, Green proposed that inhibitory processes, similar to those used during other inhibition tasks, are involved in successful language use by bilingual individuals. This inhibition of schema within the nonselected language should require executive input via the frontal and subcortical control circuitry. Indeed some f MRI and PET studies have implicated these neural regions during L2 naming and switching tasks. For example, Hernandez et al. (2000, 2001) found increased dorsolateral prefrontal cortex activation during language switching (e.g., naming an object in English or Spanish depending on a cue) but not nonswitching tasks (e.g., naming objects in just English or just Spanish). Interestingly, a similar pattern of competition mediation has been found for translation (e.g., Price, Green, & von Studnitz, 1999). In summary, bilingual speakers appear to have language lateralized within the left or dominant hemisphere, with minimal differences in L1 and L2 localization, particularly for more proficient or early L2 learners. The unique neural circuitry of bilingual speakers as well as the possible involvement of cognitive functions in mediating acquisition and use of two languages suggest that following an insult to the left hemisphere, some distinctive language signs and recovery profiles may be observed in bilingual versus monolingual aphasia cases. Whereas neurolinguistic studies to date have informed our understanding of language representation in the bilingual brain, they have not yet elucidated how the languages or their subcomponents interact. Research in the field of psycholinguistics, however, has offered some explanations of these interactions.

Psycholinguistic Views of Bilingualism

Psycholinguistic models of language processing attempt to explain steps or stages that contribute to language comprehension and production. For both monolingual and multilingual speakers, at least two levels of language representation have been forwarded: lexical and conceptual (Kroll & de Groot, 1997; Levelt, 1999). The lexical level includes information about word form and correct syntactic use, whereas the conceptual level includes information about words meaning in a real-world context. Various models have been proposed

to account for how this system works within the framework of a bilingual language system. One such model, as forwarded by Kirsner, Lalor, and Hird (1993), involves a single system in which all words related in form and meaning from both languages are stored together. In this view, the bilingual lexical system is organized similar to that of a monolingual system: Words in a bilingual speakers lexicon are organized according to morphology, not according to language. Therefore, for English-Italian speakers, words such as motivation and its Italian equivalent motivazione would be stored together, whereas words that share meaning but not morphology, such as cat and gatto, would be stored separately. Similarly, according to the Bilingual Interactive Activation model, words are interconnected, and lexical access is not language selective, as the lexicon has words from both languages (Dijkstra & van Heuven, 1998). In contrast, several hierarchical models have been proposed that specify separate lexicons with a shared concept store (e.g., Kroll & Stewart, 1990, 1994; Potter, So, Von Eckardt, & Feldman, 1984). These models differ in terms of connections between lexical stores, conceptual links, or both. For example, within the Word Association Model, L2 words are accessed through L1 (Potter et al., 1984); thus, the L2 and L1 lexicons are linked, but only the L1 lexicon has direct access to concepts via a conceptual link. Conversely, the Concept Mediation Model allows each lexicon, L1 and L2, direct access to the conceptual store (i.e., two separate conceptual links) rather than a connection between the two lexicons. Both models have received empirical support, suggesting that under different lexical retrieval circumstances, a different model may apply. To illustrate, whereas data from picture naming (Caramazza & Brones, 1980; Dufour & Kroll, 1995; Potter et al., 1984) and priming tasks (Chen & Ng, 1989; de Groot & Nas, 1991) support the Concept Mediation Model for at least bilingual speakers who are proficient in both L1 and L2, translation task data suggest that the Word Association Model better accounts for the performance of bilingual speakers who are less proficient in L2 (Chen & Leung, 1989; Kroll & Curley, 1988). Specifically, less proficient bilingual speakers more quickly translate words than name pictures in their L2, suggesting mediation through L1 for picture naming rather than a direct L2-conceptual link. Proficient bilingual speakers, however, perform picture naming and translation tasks at roughly similar speeds, supporting access to a direct L2-conceptual link. The Revised Hierarchical Model (Kroll & Stewart, 1994) addresses this lack of parsimony, allowing each language direct access to the concepts store, as well as a direct connection between the two lexical stores. Initially, L2 learners would access L2 word meanings through L1 via the connection between the L1 and L2 lexicons. With increased fluency, however, speakers are able to conceptually mediate L2 directly as a link between the concepts store and the L2 lexicon develops; importantly, the lexical link between L2 and L1 does not disappear with the formation of this new conceptual link. This model also allows asymmetry in the strength of lexical-to-conceptual connections. For example, initial dependence on L1 for L2 will result in a stronger lexical link

302 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

along the L2 to L1 path, and because L1 initially holds privileged access to meaning, a stronger link between L1 and the concepts store compared to the link between L2 and the concepts store. Indeed, this configuration has received support from investigation of the bilingual lexical semantic system via neuroimaging (Hernandez et al., 2001), pictureword interference (Hermans, Bongaerts, de Bot, & Schreuder, 1998), and semantic comparison protocols (Dufour & Kroll, 1995; Edmonds & Kiran, 2004). Influential Linguistic Variables In addition to language fluency as described above, a number of linguistic variables may influence language organization and abilities in bilingual individuals. For example, using a model similar to the Revised Hierarchical Model, Paivio, Clark, and Lambert (1988) found that concreteness affects lexical retrieval in bilingual individuals. According to the dual coding theory (Paivio & Desrochers, 1980), separate but connected verbal representations exist for each language, and these are connected to a shared imagery system. Following this theory, concrete words could be stored and accessed through a combination of verbal and image associations, whereas abstract concepts could be accessed solely through verbal coding. Therefore, in bilingual speakers, concrete words could be translated directly between the two verbal systems or indirectly through the imagery system. Indeed, Paivio et al. (1988) found that bilingual individuals demonstrated better recall of concrete versus abstract words. Similarly, Kiran and Tuchtenhagen (2005) examined the effect of imageability on the naming and semantic judgment abilities of 16 Spanish-English participants, one of whom had aphasia, and found that regardless of the presence of aphasia, task performances were better when concrete stimuli were used. Cognate status also appears influential. Generally, cognates are words that share meaning and form across languages (e.g., lamp in English and lampara in Spanish; Goral et al., 2006; Kroll & DeGroot, 1997). Across studies, however, researchers have applied different definitions and theories of cognates. For instance, for a pair of words to be considered cognates, Roberts and Deslauriers (1999) used the combination of a subjective judgment across three authors and an analysis of features and phonemic similarity with a minimum feature overlap of 70%; Lalor and Kirsner (2001), on the other hand, defined cognates as morphologically related words. These varying definitions reflect, at least in part, two competing explanations of cognates: phonological versus morphological models. According to phonological models, cognates share a phonological-sublexical base so that during language production, there is phonological activation of not only the target word but also the translated form, strengthening both forms (Costa, Santesteban, & Cao, 2005). Given this phonological overlap, cognate effects might be viewed as products of neighborhood density (i.e., the number of phonologically similar words that can be formed by changing one sound in the target word). Compared to words with few neighbors, words with dense neighborhoods, which would include their cognates, can reach higher activation levels because, as phonological segments of words are activated, they feed activation back to the target lexical node so that all words in the

neighborhood are activated. According to morphological cognate models, bilingual and monolingual systems work similarly (Kirsner et al., 1993; Lalor & Kirsner, 2001): Words are clustered together according to shared morphology, regardless of whether they belong to the same or a different language. Consistent with this view, cognates show a longterm repetition priming effect (i.e., improved accuracy in word production after recent exposure to the word) in bilingual speakers (Cristoffanini, Kirsner, & Milech, 1986). In other words, cognates produce a priming effect similar to that found with morphologically related words in monolingual speakers (Napps & Fowler, 1987). Fortunately, studies have thus far shown a positive relationship across the various definitions of cognates. For example, several investigators have found that in bilingual speakers, including those with aphasia, cognate status positively affects word retrieval and translation accuracy and speed (Goral et al., 2006; Kohnert, 2004; Lalor & Kirsner, 2001). Variation in how the term cognate is defined may result in conflicting research outcomes in the future, however. Influential Cognitive Variables As previously discussed with respect to neurolinguistic models, the languages of bilingual individuals share a more or less common neuronal system. Additionally, according to psycholinguistic perspectives, these languages have separate but connected lexicons, with a shared concept store. If there is such neural overlap, yet possibly different language stores, how are the languages kept separate, and relatedly, how is the correct language retrieved? To answer these questions, several researchers have begun to explore the role of cognitive processes. Some theorists have proposed that successful language use in bilingual individuals may be achieved through a combination of activation and inhibition. According to Paradiss (2004) Activation Threshold Hypothesis, activation of any linguistic property (e.g., word, syntactic construction, phonotactic schema) within one language causes automatic inhibition of the other. This inhibition is not language-specific. Instead the targeted item must receive more activation than competing forms, including alternate language possibilities. To inhibit these competing forms, their activation thresholds are raised. The more often a form is activated, the lower its threshold becomes, resulting in easier activation over time. Decreased use of other forms increases their activation thresholds, making them more difficult to access over time. Alternately, in a somewhat modified version of Paradiss model, Costa and Santesteban (2004) proposed that with increasing proficiency, bilingual speakers move away from inhibitory control to language-specific selection. Highly proficient bilingual speakers retrieve target words via a language-specific lexical selection mechanism rather than through inhibitory control because words from the non-target language do not compete for lexical selection and therefore do not require suppression. Importantly, both theories rely on inhibitory forces to some degree for correct lexical retrieval. Whether these inhibitory processes overlap with those used during other, nonlinguistic inhibition tasks, as suggested by Green (2005), has yet to be determined. To review, psycholinguistic research suggests that bilingual speakers have a shared, common conceptual store for
Lorenzen & Murray: Bilingual Aphasia Review

303

their languages and identifies several variables that can influence lexical retrieval accuracy and speed. Although further research is needed to resolve issues such as the nature and relationships of the lexical stores and the mechanisms underlying cognate effects on lexical retrieval, data accrued thus far through the development of both psycholinguistic and neurolinguistic models collectively support that compared to monolingual speakers, bilingual speakers develop and utilize some unique language mechanisms. Indeed, the following section reviews how bilingual clients with aphasia can experience language recovery differently than monolingual speakers.

Bilingualism and Aphasia

Bilingual clients with aphasia demonstrate a variety of recovery patterns in terms of the relative impairments of their two languages (Paradis, 1977). These patterns can be quite distinct from those of monolingual speakers and in the research literature have been described or explained in terms of output abilities, translation abilities, and cognitive factors. Because most bilingual aphasia research has focused on expressive output rather than comprehension, the primary focus of this section is recovery of expressive abilities.

(2001a) assessed and compared the recovery profiles of 20 right-handed, bilingual clients with aphasia, all of whom had learned their L2 between 5 and 7 years of age. He found that none of the following variables could reliably predict recovery patterns: (a) language status (i.e., native language or currently most used language), (b) lesion type or site, (c) environments in which the languages were used, (d) aphasia type, or (e) manner in which the languages were learned. After an extensive review of patients and their various recovery patterns, Paradis (1977) concluded that whereas many factors have been proposed to influence recovery pattern, such as age, proficiency, context of acquisition, and type of bilingualism, recovery patterns most likely result from a combination of these factors.

Recovery Patterns: Translation Skills

Bilingual clients with aphasia can demonstrate different impairment and recovery of translation abilities. Table 2 shows the four primary types of translation deficits that have currently been identifiedinability to translate, spontaneous translation, translation without comprehension, and paradoxical translationand their characteristics (Fabbro & Paradis, 1995; Paradis et al., 1982). A translation disorder may affect either language compromising translation from L1 to L2 or from L2 to L1. To account for the variety of translation disorders, Paradis (1984) suggested that translation may consist of two components, one that allows translation from language A to language B, and another component for translation from language B to language A. Following this, when a translation disorder occurs, it may be due to (a) one of the components being negatively affected, with the other part intact, (b) both components negatively affected to similar degrees, or (c) both components negatively affected but to different degrees. The ability to translate is a cognitive task that goes beyond being able to speak and understand two languages (Paradis et al., 1982). The skill of translating involves switching languages and is normally under voluntary control; when there is a translation disorder, however, the translation may be automatic, such as in the case of spontaneous translation (Perecman, 1984). Perecman differentiated spontaneous translation from language mixing in terms of their different levels of cognitive processing: She proposed that a

Recovery Patterns: Output Abilities

During the acute recovery phase (i.e., up to approximately 4 weeks after aphasia onset), a client may experience substantial change across either or both languages, dramatic differences between the languages, or both, resulting in a variety of dynamic, bilingual aphasia recovery patterns (see Table 1; Fabbro, 2001a; Green, 2005; Paradis, 1977). For instance, Fabbro (2001a) examined 20 right-handed, bilingual ItalianFriulian speakers using the Bilingual Aphasia Test (BAT; Paradis & Libben, 1987) and found that approximately 65% of these clients exhibited parallel recovery, 20% had greater impairment of L2, and 15% had greater impairment of L1. Similarly, in a review of 132 cases, Paradis (2001) found that approximately 61% exhibited parallel recovery, 18% differential recovery, 7% blended, and 5% selective. Despite agreement that diverse recovery patterns may occur, exactly what contributes to a given clients recovery pattern is still debated. Over the course of several years, Fabbro
TABLE 1. Recovery patterns found in bilingual aphasia. Recovery pattern Parallel recovery Differential recovery Antagonistic recovery Alternating antagonism Blending recovery Selective aphasia Successive recovery Note.

Language characteristics Recovery of languages parallels the premorbid relative abilities. If one language were stronger premorbidly, it would return to being stronger. One language is recovered much better than the other compared to premorbid abilities. One language is initially available, and as the other language recovers, the initially available language disappears. Repetition of the above pattern with languages alternating in availability. This may occur within cycles ranging from 24 hr to several months. Uncontrollable mixing of words and grammatical constructions of two or more languages even when attempting to speak in only one language. This should not be confused with the common bilingual practice of code switching. Language loss only in one language with no measurable deficit in the other. The recovery of one language before the other(s).

Adapted from Paradis (2004) and Fabbro (2001a).

304 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

TABLE 2. Translation deficits found in bilingual aphasia. Translation deficits Inability to translate Paradoxical translation Translation without comprehension Spontaneous translation Note. Language characteristics Inability to translate from either language to either language Ability to translate in one language but not the other Ability to translate language promptly but lack the ability to understand its meaning Inability to inhibit translating, producing involuntary translations of utterances that they or others have said

Adapted from Fabbro (2001b).

spontaneous translation deficit may result from a prelinguistic processing disorder, whereas language mixing is a linguistic level disorder. Accordingly, a translation disorder may be part of a larger conceptual disorder that affects other behaviors in addition to language. It cannot be assumed, however, that the presence of spontaneous translation in the speech of a person with aphasia indicates a deficit. Grosjean (1985) proposed that some of what appears as involuntary translation may in fact be a deliberate communicative strategy used to increase communication effectiveness.

Recovery Patterns: Impact of Concomitant Cognitive Deficits

Given that cognitive abilities have been implicated in the development and use of more than one language (e.g., Green, 2005), it is possible that cognitive deficits could contribute to at least some bilingual recovery patterns. For example, impaired cognitive control could lead to differing degrees of too much or too little inhibition, and consequently, impaired ability to maintain a given language and various recovery patterns: permanent inhibition (i.e., selective recovery), temporary inhibition (i.e., sequential recovery), alternating inhibition (i.e., antagonistic recovery), greater inhibition in one language (i.e., differential recovery), and loss of inhibition (i.e., blending; Green, 1986). A possible result of impaired top-down control is pathological code switching, a pragmatic disorder in which individuals are unable to control language switching and consequently code switch in unacceptable circumstances leading to communication breakdowns (Fabbro, Skrap, & Aglioti, 2000). That is, due to impaired top-down attentional control, access to the appropriate set of lexical concepts within the target language may be compromised by competition from the other language. For example, Fabbro et al. reported on a client who exhibited pathological switching even when the interlocutor did not understand the alternate language. In addition to the pathological switching, the client exhibited behavioral and verbal disinhibition. Muoz and colleagues (1999) also observed that only participants with aphasia demonstrated higher frequencies of certain switching patterns, communication difficulties due to code switching when the social or pragmatic expectation was continued use of one language, frustration with an inability to use the appropriate language, or an inability to stop pathological code switching. Similarly, Marin, Abutalebi, Engelborghs, and De Deyn (2005) reported on uncontrollable language mixing following deep left frontal lobe lesions, suggesting the

involvement of a control mechanism. In a study involving a trilingual speaker of English, French, and Hebrew, Goral et al. (2006) found that involuntary switching (referred to as language interference) occurred mainly in French, the least recovered language, and that the interference occurred between the languages that shared the most vocabulary. For example, target English words were more frequently substituted with French rather than Hebrew words, despite the clients stronger Hebrew. As discussed with spontaneous translation, the presence of code switching does not only imply an inability to control language output: Whereas increased mixing may at times be atypical according to normal standards, for clients with aphasia it could be a conscious or unconscious self-cuing strategy to access the correct word in either language or to enhance communication in general (Grosjean, 1989). In this case, the language switch or mix could function as a paraphasia that only interferes with communication if the interlocutor does not speak the language.

Recovery Patterns: Learning Factors

Learning mechanisms such as Hebbian learning may also contribute to variation in bilingual aphasia recovery (Green, 2005). Hebbian learning, a primary means of function restitution, involves increased synaptic efficacy through neuronal reconnection, resulting in increasingly strong connections over time (Hebb, 1949). Stronger undamaged neuronal circuits compete to control output, suppressing not only circuits temporarily malfunctioning due to diaschisis but also the repair and resurgence of damaged circuits. Selective, antagonistic, and differential recovery may therefore reflect restitution via Hebbian learning in which only one pathway was strengthened (Green, 2005): (a) selective recovery may be due to progressive use of one language, which consolidates and isolates its network from the other; (b) continued growth in dominance of one language could eventually completely inhibit use of the other language, resulting in an antagonistic recovery; and (c) if the lexical semantic system of one language is impaired marginally more than that of the other, that weaker language may not benefit from Hebbian learning, resulting in differential recovery of the already stronger language.

Recovery Across Language Processes and Modalities

As with monolingual aphasia, recovery varies significantly across language processes and modalities with bilingual aphasia.

Lorenzen & Murray: Bilingual Aphasia Review

305

This section examines pertinent research regarding recovery of syntax, semantics, reading, and writing. Other language processes (e.g., phonology, pragmatics) and modalities (i.e., gesture) have received less empirical attention, and thus further research is needed to determine whether bilingual clients with aphasia demonstrate unique recovery patterns in these aspects of communication. Morphosyntax. Patterns of grammatical recovery can be highly variable in bilingual speakers due to core linguistic differences in their languages, such as the relative importance of specific linguistic properties, for example, morphological markers, word order, and prosody (Green, 2005). As explained by Paradis (1988), language impairments depend on how the system can break down, which is determined by the structure of the language system. For example, syntactic disorders apparent at the surface of a speakers grammar are dependent on the languages structure; accordingly, morphological breakdown may be more noticeable in a highly inflected (e.g., Spanish) versus minimally inflected language (e.g., English). Indeed, when evaluating agrammatic ItalianFriulian speakers who were exhibiting parallel recovery, Fabbro (2001a) observed different levels of pronoun omission in Italian versus Friulian. Italian is a partial pro-drop language, meaning it allows speakers to drop the subject pronoun if the subject can be inferred from the contextfor example, if an Italian speaker stated, Mio padre avvocato (My father is a lawyer), his or her next sentence could be Lavora in centro (Works downtown) in which the pronoun he can be excluded; conversely, Friulian, like English, is a non-pro-drop language, meaning speakers must use the subject pronoun regardless of the referents transparency. The bilingual speakers with agrammatism in Fabbros study made more frequent pronoun deletion errors in Friulian than in Italian, suggesting that this error pattern was most likely a consequence of required pronoun use in Friulian. Relatedly, because Italian allows pronoun dropping, pronoun omission errors were nominal: The speakers with aphasia appeared to make fewer errors, simply because they were able to avoid this syntactic mistake. Another example pertains to case- and gender-marking deficits, which vary subsequent to aphasia onset depending on the individuals languages (Menn, OConnor, Obler, & Holland, 1995). Speakers with aphasia who use languages with case marking (e.g., German, most Slavic languages) often demonstrate substitution errors in case marking (e.g., use an incorrect morpheme to indicate nominative/subjective case). As case marking is limited to the possessive s in English, case errors are infrequent in English speakers with aphasia. Similarly, gendermarking errors will be common if speakers with aphasia use languages that have gendered nouns (e.g., for the English noun phrase a /the house, Spanish requires the use of a gender-specific determiner una/la and the feminine word ending a). Variable syntax profiles among bilingual speakers with aphasia may also result from differences in the nature of their syntactic impairment. There are two general views to account for syntactic problems in aphasia: a central deficit disorder, in which syntax competence or knowledge is impaired, and a peripheral deficit disorder, in which syntax performance is impaired (Grodzinsky, 2000; Paradis, 1988). Impaired syntax

competence assumes that all grammatical knowledge is centrally represented, and thus, an impairment would affect all language modalities. Conversely, a peripheral deficit is defined as a modality-specific deficit and is considered a performance deficit. A peripheral deficit could, then, produce a double dissociation wherein an individual with aphasia may, for example, demonstrate syntactic breakdowns in writing but not speech or reading (Bub & Kertesz, 1982; Paradis, 1988). With respect to bilingual speakers with aphasia, similar deficits across languages would suggest a central deficit (Paradis, 1988). If the deficit is peripheral, there may be a dissociation between languages, depending on what syntactic operation has been affected, resulting in differential language outcomes such as: (a) if there is a structure dependent deficit (e.g., difficulties with tense or case marking), the language with the more extensive morphological system will be more impaired; (b) if lexical access is impaired, the language with greater use of free or bound grammatical morphemes will be more impaired; and (c) if a syntactic level deficit occurs, the language that relies more heavily on word order would be more impaired. As proposed by MacWhinney, Bates, and Kiegl (1984), cue strength may also account for differences in productive syntactic deficits among speakers various languages. These researchers demonstrated that cue strength, the degree to which a given grammatical aspect can be utilized for sentence interpretation, varies across languages: The stronger the relative cue, the more likely the cue will be preserved. To illustrate, participant-verb-object is the canonical order for both Spanish and English, but Spanish speakers have more freedom to vary word order from this canonical order. Therefore, canonical word order has greater cue strength in English, and accordingly, its use often remains preserved following aphasia onset in English speakers. In languages in which canonical word order has a lower cue strength, such as Spanish, it may be less preserved. Similarly, to compare grammatical errors across languages with differing levels of morphology and word-order freedom, Bates, Friederici, and Wulfeck (1987) examined adults with Brocas aphasia, Wernickes aphasia, or no aphasia who spoke English, Italian, or German. All three languages have a canonical, participant-verb-object order, but Italian and German also allow quite flexible word order. To compensate for this flexibility, these two languages utilize morphology for sentence interpretation in that (a) German has a significant and productive case system attached to articles, and (b) Italian requires reliance on agreement markers. Conversely, English utilizes little morphology, necessitating dependence on word order for interpretation. These language differences, according to a traditional impairment perspective, could place German and Italian language users at a disadvantage if aphasia impairs morphology more than other language features (e.g., word order). Bates et al. found, however, that both Italian and German participants with Brocas aphasia maintained morphology use to a much greater degree than would be expected; in fact, no aphasia type was associated with omission rates equal to that demonstrated by English speakers with aphasia. Although similar syntactic recovery patterns have been observed in other morphologically rich languages such as Spanish (e.g., Ostrosky-Solis, Marcos-Ortega, Ardila,

306 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

Roselli, & Palacios, 1999), additional studies comparing the relative maintenance of morphology across languages would be beneficial, with tighter control for language impairment and testing tools. Regardless, this line of research indicates that reconsidering previous assumptions about morphosyntax and aphasia is needed, particularly given that these languagespecific differences influence the appropriate identification and selection of morphosyntactic treatment targets. Another example relates to the proposed dichotomy between agrammatism and paragrammatism. Agrammatism is typically characterized by telegraphic speech, including omission of grammatical inflection and function words, reduced phrase length and syntactic complexity, and Brocas area lesions (Goodglass, Kaplan, & Barresi, 2001); in contrast, paragrammatism is typified by empty speech (i.e., words or phrases meaningless to the communication situation, including indefinite terms, neologisms, other paraphasias, and repetitions), inflection substitution errors, frequent use of functors, and Wernickes area lesions. This dichotomy of omission versus substitution, however, may be overstated due to the disproportionate number of aphasia studies done in English, a word-order strict, morpheme-impoverished language. For example, in a study of agrammatism in two bilingual Spanish-Catalan speakers with Brocas aphasia, the participants were observed to make verb errors in both languages that were considered uncharacteristic of traditional descriptions of Brocas aphasia (de Diego Balaguer, Costa, Sebastin-Galles, Juncadella, & Caramazza, 2004). Therefore in bilingual speakers, patterns of grammatical deficits in clients with aphasia can differ from what was previously assumed, and whether and how these traditional grammatical profiles apply to other languages, particularly within bilingual speakers, requires further examination. A growing literature thus indicates that the linguistic characteristics of each language used by bilingual speakers with aphasia must be considered when describing their morphosyntactic strengths and weaknesses. Failure to consider these influential linguistic factors may result in inaccurate documentation of the presence, severity, and nature of grammatical deficits. Lexical semantics. Specific considerations also apply to the recovery of lexical semantic abilities in bilingual speakers with aphasia. Whereas much research has focused on monolingual speakers of English, research involving monolingual speakers who speak a language other than English suggests that the types and frequencies of paraphasias vary across languages. Consequently, types and frequencies of paraphasias may also vary across a bilingual speakers different languages. For example, Ardila (2001) observed that Spanish speakers produce more phonemic or literal paraphasias involving vowels than do speakers of other languages. This is possibly due to a one-to-one correspondence of vowels and their sounds, making vowels in the Spanish language more salient. Research with bilingual speakers has also highlighted some unique paraphasia considerations. For instance, semantic paraphasias may sometimes result from interlanguage interference, with the speaker choosing the correct word, but from the wrong language. Goral et al. (2006) reported on a trilingual speaker whose L1 was Hebrew but who used English and French more frequently than Hebrew at the time

of his aphasia onset. Levels of interlanguage interference errors varied across his languages, which the authors attributed to differing levels of lexical connectedness among his languages. That is, when the client was speaking in French, he produced many non-target language words in English, which often were cognates. Conversely, when speaking Hebrew, a language less related to either French or English, he produced significantly fewer interference errors. Therefore, interlanguage interference represents a possible source of paraphasias not encountered when dealing with aphasia in monolingual speakers. Several unique factors may influence lexical access speed, accuracy, and recovery in bilingual speakers with aphasia. For example, investigators have begun to explore the effects of cognate status on naming skills and have found that cognates elicit better pretreatment performance than noncognates in participants with aphasia (Kohnert, 2004; Lalor & Kirsner, 2001; Roberts & Deslauriers, 1999). This benefit may not, however, apply to both languages, as Roberts and Deslauriers (1999) found that in adults with aphasia, cognate status only improved L2 naming accuracy. Therefore, this line of bilingual aphasia research, albeit limited, has consistently revealed improved naming accuracy and speed with cognates versus noncognates, with additional research needed to identify the limits of this generalization. Reading and writing. Recovery of reading and writing abilities in bilingual individuals with aphasia has received some empirical attention. Some research suggests that these individuals may utilize different reading strategies in their different languages, particularly if their languages diverge in terms of orthography transparency. For instance, with English, a language with nontransparent orthography (i.e., one cannot always sound out words), individuals may use a combination of phonological and logographic (i.e., whole word) reading approaches (Sampson, 1985). Spanish, in contrast, has transparent orthography, as words are pronounced as written, and allows accurate reading aloud even when words are completely unknown (Ardila, Rosselli, & Pinzon, 1989). As a result, individuals may more easily and regularly employ a phonological route to read Spanish (Ardila, 1991). Iribarren, Jarema, and Lecours (1999) have pointed out, however, that whereas transparent orthography may favor a phonological reading route, it does not preclude the use of logographic reading. Indeed, researchers have identified the use of logographic reading in participants from language backgrounds with transparent orthographies (e.g., Basso & Paulin, 2003; Iribarren et al., 1999). Clinicians should be aware of these potential cross-linguistic differences in reading routes, which may in turn produce different patterns of reading deficits subsequent to aphasia onset in bilingual individuals. For example, disturbance of just one reading route could impair reading one language more than the other or, at the very least, influence what type of stimuli might be appropriate. Reading disturbances subsequent to aphasia onset may include surface dyslexia, deep dyslexia, or phonological dyslexia (Drouesn & Beauvois, 1979; Marshall & Newcombe, 1973). Just as the surface manifestations of these dyslexia types differ, so may their manifestations across languages, making the identification of such deficit patterns difficult. For
Lorenzen & Murray: Bilingual Aphasia Review

307

example, surface dyslexia, characterized by a regularization and thus misunderstanding of irregularly spelled words, may be uncommon in speakers of orthographically transparent languages (Iribarren, 2007). Conversely, compared to clients who regularly employ both routes to read their language(s), clients using languages that primarily rely upon a phonological reading route may be at greater risk for severe forms of phonological dyslexia, a reading disorder distinguished by impaired use of the phonological reading route. As with reading, orthography transparency may influence the types of writing profiles observed in bilingual clients with aphasia. For example, Spanish, Italian, and Russian have a phonological writing system in that words are spelled as they sound (Ardila, Rosselli, & Ostrosky-Solis, 1996); that is, the letters are pronounced the same way, regardless of the word in which they are found. English and French, conversely, are not purely phonological, as evidenced by the fact that a written word in either of these languages may be quite different from its phonographic form; as a result, writing in these languages is achieved through a combination of grapheme/ phoneme correspondence (i.e., a phonological route) and visual gestalt, or whole-word recognition (i.e., lexical semantic route). Therefore, use of either a phonological and/or a lexical semantic system for writing will depend, at least in part, on the particular linguistic characteristics of the language. Ardila et al. (1996) argued that in languages such as Spanish, two different types of writing errors can be found: homophone or orthographic errors and nonhomophone errors. Homophone errors are phonologically correct and occur when two letters have the same phonological quality, could be interchanged, and thus still produce a word that sounds correct, despite being misspelled. For example, in Spanish, letters g and j both are associated with the /x/ sound, so mujer (woman) could be spelled phonetically as either muger or mujer. These spelling errors exist in healthy populations. Conversely, nonhomophone errors involve some letter additions, omissions, or substitutions, which change the words written representation (e.g., muer for mujer). These errors are only found in brain-damaged individuals. For instance, using a dictation task, Ardila et al. examined the occurrence of homophone and nonhomophone errors in the written Spanish of 92 healthy participants and 14 who had brain damage (i.e., individuals with Brocas aphasia, Wernickes aphasia, or right hemisphere damage). Whereas all groups made homophone errors, the groups with aphasia most frequently made nonhomophone errors such as letter substitutions, additions, and omissions. Additionally participants with Brocas aphasia made semantic paragraphias (e.g., lpiz Y papel, judicial Y policia), and participants with Wernickes aphasia made more morpheme substitutions and fewer omissions than those with Brocas aphasia. Due to increased transparency in the orthography of Spanish and other similar languages (e.g., Italian and Russian), it may be possible to identify unique writing errors such as homographic and nonhomographic, with the latter being prevalent only among the brain injured population. The presence of these errors and the subtypes made (e.g., letter additions vs. omissions) can vary across aphasia type. To date, most reading and writing research that extends beyond English, however,

has been done with monolingual participants with aphasia. Accordingly, future studies are needed to examine whether and how these different surface manifestations identified in monolingual individuals might vary when a person reads and writes in more than one language, particularly if the languages differ in terms of orthography transparency. In summary, although bilingual aphasia research has revealed significant variation in recovery across morphosyntax, lexical semantics, reading, and writing, additional investigations are needed to understand fully the nature of this recovery process. Regardless, unique recovery patterns dictate that novel assessment and treatment approaches, some of which are reviewed below, should be considered when providing direct services to patients with bilingual aphasia.

Implications for Assessing and Treating Clients With Aphasia

Due to the aforementioned differences between bilingual versus monolingual aphasia, managing bilingual aphasia will necessarily diverge, at least in part, from monolingual aphasia diagnostic and treatment practices. That is, traditional procedures documented as effective with monolingual speakers may require alteration or complete revision to meet the needs of bilingual speakers. Although nominal empirical research is currently available to guide effective management of bilingual aphasia, growing anecdotal evidence is emerging from clinical practice. Because bilingualism varies dramatically across individuals, it is imperative that clinicians and researchers avoid several common false assumptions. First, it cannot be assumed that premorbid abilities in each language were equivalent, or that being bilingual implies a certain level of proficiency for all bilingual speakers (Muoz & Marquardt, 2003). Second, SLPs should not attribute postmorbid differences between the languages to brain damage, as they may already have differed premorbidly (Kiran & Tuchtenhagen, 2005). Following this reasoning, SLPs must consider the various domains of language use when developing management plans for their bilingual clients with aphasia.

Assessment
To obtain an accurate assessment of language proficiency in bilingual clientele, several measures must be taken, some of which overlap with those used to evaluate monolingual clients and some of which are unique to meet bilingual clients needs. When dealing with the sudden onset of aphasia, such as in stroke, there are certain recommendations for approaching assessment as recovery unfolds. During the acute recovery phase, as is recommended for monolingual speakers (e.g., Holland & Fridriksson, 2001), assessment should focus on evaluating basic communication needs due to the dynamic nature of acute recovery (Fabbro, 2001a). During more chronic recovery phases when language abilities stabilize, more comprehensive assessment can be completed. Furthermore, during and following treatment, periodic reassessments are required to monitor for changes in recovery patterns. Throughout this assessment process, SLPs must also

308 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

dedicate time to counseling and educating caregivers and family members who may be confused and alarmed by the sudden onset of aphasia, other concomitant signs, or recovery patterns. When planning an assessment for a bilingual client, several important areas should be considered to identify specific linguistic and possibly cultural aspects that may affect both evaluation and intervention. An essential first step is collecting a history of language use from the client, family members, or others who may be familiar with the clients language use (Muoz & Marquardt, 2003; Muoz et al., 1999; Roberts & Deslauriers, 1999). Research indicates that self-reporting, accrued through interviewing, self-rating scales, or language use questionnaires, is generally consistent with actual linguistic ability and thus an efficient and effective approach to obtain language histories (MacIntyre, Noels, & Clement, 1997; Ross, 1998). For example, the language use questionnaire developed by Muoz et al. (1999) includes questions about language acquisition, educational history, and language use, and has been used with both clinical and nonclinical populations. More recently, Marian, Blumenfeld, and Kaushanskaya (2007) developed the Language Experience and Proficiency Questionnaire (LEAP-Q), which covers acquisition history and context, language use, language preference, and a proficiency rating, and has validity and reliability data available for a nonclinical population. The authors, however, emphasized that LEAP-Q is not yet intended for a clinical population, even though it may be useful in obtaining language histories from these individuals. If SLPs opt to create their own questionnaires, they should include questions pertaining to the following: age and manner of acquisition (i.e., home vs. formal schooling), current use of each language, relative proficiencies, literacy levels in each language, and modalities in which each language is used (Paradis, 2004). Identifying bilingual clients premorbid proficiency levels is germane to establishing the severity of impairment within each language and across modalities. For example, during informal interaction, a client may demonstrate only slightly stronger language abilities in one language than the other, suggesting a parallel deficit. It may be the case, however, that the client was significantly more proficient premorbidly in one of the languages and thus has sustained a significant loss in one language and only minimal in the other. Conversely, as not all bilingual speakers are literate in both languages, an apparent reading deficit may actually reflect a premorbid state. When formally assessing the clients language abilities, it is imperative that each language, whether used regularly or only for specific purposes, be thoroughly evaluated, using tests designed for bilingual speakers (Paradis, 2004). Assessment of both languages is important for several reasons. First, clients may be unaware of which of their languages is more impaired and will make life choices based on their perceived deficits. Second, if the SLP only speaks one of the languages, he or she may wrongly perceive or assume that the other language is less impaired when it is not. A complete assessment will ensure an accurate picture of the strengths and weaknesses within each language. It may also reveal language deficits that are only detectable in one of the languages due to structural differences of the languages. The

goal is to understand relative deficits across languages and modalities so as to determine which languages to include in therapy or, alternately, to identify specific goals for each language. If the SLP does not speak the clients language, interpreter assistance may be needed (see ASHA, 1989, for guidelines on proficiency requirements). Due to difficulties associated with using interpreters (e.g., training, reliability), interpreters should only be employed when no SLP is available who can speak the clients language (ASHA, 1989). Importantly, family members should only be used as a last resort, because it may be difficult for them to participate objectively in assessment without providing personal input into the process. Whereas one might assume that an SLP who is proficient in each of the clients languages would be best qualified to administer these tests, this is not always the case. For example, if a client is suspected of pathological switching, it is important that the SLPs doing each language assessment speak only that language (Grosjean, 1998). For instance, when assessing a clients Spanish skills, it would be ideal to have a tester who only spoke Spanish so that any English utterances could not be assumed to be intentional. With respect to formal test selection, Paradis (2004) warned against the folly of using tests that are simply translations of English tests; such tests are constructed with little or no consideration of linguistic and possible cultural differences. Instead, SLPs must identify linguistically equivalent tests that evaluate similar types and levels of abilities in each of their clients languages. Linguistic equivalence implies that all areas of linguistic comparison (e.g., vocabulary, syntax structures) are of equal difficulty in the respective languages. For example, across tests, grammatical constructions must have the same level of difficulty or complexity; simply translating a particular grammar tense or construction does not guarantee that the construction is equally difficult in the other language. Minimal pairs or rhyming words that are used for phonological tasks will obviously differ if simply translated and thus no longer meet the phonological tasks requirements. Likewise, whereas vocabulary on aphasia batteries is typically controlled for frequency of use, translation equivalents in another language may have different frequencies, making them more or less common or difficult. Finally, because languages differ structurally (e.g., word order, inflectional morphology, complexity of construction such as passive voice), the structural complexity and relevance of a structure may also vary across languages. To be equivalent, a structure of similar complexity rather than similar use needs to be chosen. These are only a few ways in which structures may differ, exemplifying why an assessment battery needs to be completely rewritten prior to using it cross-linguistically. As a result, it may be most wise to use a testing tool that has been specifically designed for multiple languages. Perhaps the most frequently used bilingual aphasia test, at least in the research literature, is the BAT (Paradis & Libben, 1987). The BAT does not classify aphasia type but rather measures bilingual clients ability to use each of their languages in unilingual settings. That is, it identifies which language skills and linguistic structures have been affected
Lorenzen & Murray: Bilingual Aphasia Review

309

for each language. The BAT includes a detailed language use history questionnaire and sections to assess deficits in language use and translation, as well as language interference issues. A computer program is available to evaluate responses in more than 100 different languages and make comparisons across languages. Importantly, the BAT is not simply a translation of an English aphasia test; rather, tests in each language were developed with linguistically and culturally appropriate words and syntactic structures (Fabbro, 2001a). A few other aphasia batteries are available for bilingual clientele and can be used for classification of aphasia. These include the Multilingual Aphasia ExaminationSpanish Version (Benton & Hamsher, 1994; Rey, Sivan, & Benton, 1991) that has been standardized and normed on a Latino sample, the Western Aphasia BatterySpanish Version (Kertesz, Pascual-Leone, & Pascual-Leone, 1990), a Spanish version of the Psycholinguistic Assessment of Language Processing in Aphasia (Kay, Lesser, & Coltheart, 1992), a Spanish version of the Boston Diagnostic Aphasia Examination (Goodglass & Kaplan, 1986) that has available norms (Rosselli, Ardila, Florez, & Castro, 1990), and the Aachen Aphasia Test (Huber, Poeck, Weniger, & Willmes, 1983), which is available in several languages. In addition to standardized testing, SLPs often obtain and analyze language samples for quantity and quality of production. Here again, a cross-linguistically comparable measure must be used. Because of differing structural characteristics of languages, measuring language quantity according to procedures developed for English may under- or overestimate problems in other languages. For example, syntactic complexity is often quantified by computing the mean length of utterance, the number of verbs per utterance, and the number of subordinate clauses (Paradis, 2004). Because norms for these measures vary across languages, the clients performance in each language needs to be compared against its respective language norms, rather than solely compared to the other language. Additionally, Paradis suggested that analysis procedures may require modifications. Take for example the typetoken measure, which typically divides the number of different words (i.e., types) by the total number of words (i.e., tokens) to measure lexicon diversity. To remediate some problems associated with lexical structure differences across languages, Paradis suggested that when doing a typetoken analysis, only lexical items (not grammatical words) should be counted. For instance, a language without articles or copulas would have a higher typetoken ratio than a language with these word types, because it would have fewer redundant words, appearing artificially more diverse. Similarly, if a language uses noun phrases as compound words, each constituent of the compound should be counted as one word when comparing to languages in which several separate words mean the same thing. Finally, if reading and writing are assessed, SLPs must be aware of any visual field cuts or neglect that could differentially affect two languages if they are read from opposite directions (Paradis & Libben, 1987). Therefore, SLPs must be cognizant of these and other possible language differences so as to make accommodations that will ensure that bilingual clients abilities are accurately assessed. Failure to do so could result in incorrect assessment of aphasia presence, severity, or classification.

Due to increasing evidence that possible concomitant, cognitive disorders may influence aphasia signs and recovery in monolingual speakers (e.g., Murray, 1999) as well as bilingual speakers (e.g., Green, 2005), cognitive testing may be indicated. Similar to the limited availability of language tests, there are few appropriate tools for examining multilingual clients cognitive abilities, although researchers are beginning to tackle this issue (Rey et al., 2001). A limited set of translated language and cognitive tests are listed in the Appendix, and a more complete discussion of normative data available for various language and cognitive tests can be found in Roberts (2001). In summary, clinicians working with the bilingual aphasia population must consider and thus evaluate their clients from a multitude of angles to develop the most effective and efficient treatment (Fabbro, 2001a; Roberts, 1998). For example, Green (2005) advocated that an essential first step in planning treatment is to assess and understand the clients recovery pattern across and within languages to create practical goals that will target specific loci of breakdown in the language system. Next, clinicians need to understand (a) what may be driving particular recovery patterns, (b) whether cognitive deficits are in any way contributing to language behaviors, and (c) how these factors interact with their clients needs and wishes. Additionally, depending on clients needs and preferences, clinicians must determine which language or languages to target and then maximize treatment efficiency by analyzing those languages for similarities of underlying structure to determine how best to exploit these similarities in concert with natural learning processes. Finally, the clients communication community can also influence assessment and treatment choices and recovery patterns and thus be utilized as a valuable resource in all aspects of client management (Laganaro & Overton Venet, 2001).

Treatment
Whereas researchers often provide suggestions regarding treatment of bilingual aphasia (e.g., Paradis, 2004; Roberts, 2001), sufficient experiential support of these suggestions is currently lacking. Additionally, the available empirical literature has so far focused on a limited set of linguistic abilities (e.g., word retrieval; Edmonds & Kiran, 2006). These initial investigations, however, can provide both clinicians and researchers with ideas for developing treatment protocols for bilingual clients and highlight the many management issues unique to the bilingual client population that still need to be resolved in future studies. For most bilingual clients, the same principles applied to design effective therapy for monolingual speakers would be used. For example, selected treatment strategies should ensure client success by choosing accessible language targets (Roberts, 2001). Additionally, some existing traditional treatments have been found effective with bilingual clients, including the general stimulation approach (Watamori & Sasnuma, 1976, 1978)in which intensive auditory stimulation, repetition, naming, reading, and writing tasks are practiced at varying levels of linguistic complexityand phonemic cuing (Roberts, de la Riva, & Rhaume, 1997), in which initial phonemes of words are used to facilitate lexical

310 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

retrieval. Other treatment protocols (e.g., Melodic Intonation Therapy, which uses music-like intonation and rhythm to aid spoken word and phrase production; Sparks, Helm, & Albert, 1974) have been successfully used with monolingual speakers of other languages (see Roberts, 2001, for a review), but additional research is needed to determine not only the effects of these treatments on bilingual aphasia but also the degree of cross-linguistic generalization they can achieve. For example, to address word retrieval difficulties, Galvez and Hinckley (2003) provided to a client with transcortical motor aphasia a cuing hierarchy treatment in each of his languages. The client, a 71-year-old left-handed man, reported equal proficiency in Spanish and English prior to his stroke; he reportedly used both languages daily, Spanish at home and English at work. Whereas treatment produced within-language improvements across language modalities, no cross-linguistic generalization was observed, suggesting that this treatment may not generalize across languages. Aphasia treatment must also extend beyond protocols developed for monolingual English clients. Goals should be based on pre- and postmorbid language proficiency and use patterns (Roberts, 2001); for example, reading therapy should only be applied to the languages that the client premorbidly used for reading. SLPs should also consider specific characteristics of each of their clients languages, as syntactic impairments may differ across languages (Ardila, 2001; Bates et al., 1987) and different strategies may be more or less appropriate (Cuetos & Mitchell, 1988; Ostrosky-Solis et al., 1999; Roberts, 2001). For instance, bilingual Spanish/ English speakers may use gender as a cuing strategy for word retrieval in Spanish (Roberts, 2001) but not in English due to the lack of gender marking in English. Accordingly, treatment solely based on a monolingual English framework may not exploit the full range of potential strategies available. With respect to remediating specific language domains, reading and naming treatment studies have demonstrated cross-linguistic generalization when training has focused on shared aspects of clients languages (Kiran & Edmonds, 2004; Laganaro & Overton Venet, 2001). Indeed, a cognate therapy approach is based on the possibility that the two lexicons of a bilingual speaker may be functionally and neurologically interconnected, which should thus allow across- as well as within-language gains (Costa et al., 2005; Kiran & Tuchtenhagen, 2005; Lalor & Kirsner, 2001; Roberts & Deslauriers, 1999). For example, Kohnert (2004), adopting a phonological view of cognates, explored cross-linguistic generalization following a therapy to target naming of 10 cognates and 10 noncognates. The participant with aphasia, a proficient Spanish-English speaker whose L1 was Spanish, first received the cognate-based treatment in Spanish and then English. Generalization occurred from L2 to L1 for both cognates and noncognates following English treatment, but only for cognates after Spanish treatment. Therefore, a cognate effect was found with generalization for cognates in both directions. Relatedly, Laganaro and Overton Venet (2001) reported on a Spanish-English speaker with aphasia whose L1 was Spanish but who used English professionally at work. Because reading in both languages was equally impaired, these researchers concluded impairment of similar reading processes in both languages and

further hypothesized that if shared reading processes are impaired, treatment aimed at improving the shared processes should result in cross-linguistic generalization. Accordingly, they provided this client two different treatments: first, a lexical, letter-by-letter reading inhibition task, and second, a phonological assembly and encoding task, which included phonological blending activities. Indeed, the lexical treatment evoked generalization from the treated to the nontreated language, supporting the notion that targeting strategies shared by the languages promotes generalization. Which language is trained may also influence therapy efficiency. Kiran and Edmonds (2004; Edmonds & Kiran, 2006) have shown that lexical semantic treatment of participants weaker language may produce generalization to their stronger language. In their 2004 study, a lexical semantic treatment was provided to two Spanish-English speakers with equivalent, postmorbid, across-language deficits, one of whom was premorbidly English dominant, while the other was premorbidly balanced. They found that premorbid dominance contributed to cross-linguistic patterns of generalization: For the participant with a balanced Spanish-English profile, withinlanguage and cross-linguistic generalization was found. That is, treatment for this individual was initiated in Spanish, and generalization was found to semantically related Spanish words (e.g., manzana-naranja), and in English for both the trained item (e.g., manzana-apple) and semantically related items (e.g., manzana-orange). For the participant who was English dominant, within-language generalization to related semantic items was found only in the dominant language (e.g., apple-orange) but not in the weaker language (e.g., manzananaranja). Cross-linguistic generalization was only found following treatment of the weaker language (Spanish), with generalization to words in the dominant language (English) for both the trained item and semantically related items; crosslinguistic generalization to the weaker language (Spanish) did not occur following treatment of the dominant language (English). In 2006, Edmonds and Kiran again reported crosslinguistic generalization in 2 premorbidly English dominant participants only when their weaker language (Spanish) was treated. A third participant, who was premorbidly equally bilingual, was only trained in Spanish and demonstrated cross-linguistic generalization to English. Whereas additional studies encompassing more participants and directly comparing treatment protocols are needed, these preliminary findings suggest bilingual aphasia treatment can be designed to maximize cross-linguistic generalization and, thus, treatment efficiency, via training targets in only one language. Our recent review of the literature identified no empirical studies of syntax treatment for bilingual aphasia. Based on reports of cross-linguistic generalization in other language domains (e.g., Kiran & Edmonds, 2004), however, syntax treatment may be most effective for bilingual speakers when focused on underlying syntactic processes (e.g., inflectional phrase structure) shared across languages rather than on surface structures of just one language. SLPs also need to recall that different syntactic outcomes across languages may reflect the relative importance of corresponding surface structures within each language. For example, as pointed out by
Lorenzen & Murray: Bilingual Aphasia Review

311

Bates et al. (1987), the maintenance of morphosyntax may be at least partially affected by its relevance. Cognitive treatments may be appropriate for those bilingual clients with concomitant cognitive deficits. For instance, Kohnert (2004) developed a nonlinguistic cognitive treatment for a 62-year-old bilingual man with severe nonfluent aphasia. His L1 was Spanish, but he had lived in the United States for 25 years and used English in work and social settings and both languages at home. Kohnert hypothesized that improving cognitive processing efficiency to support language behavior would result in cognitive as well as language improvements. Cognitive training was provided across 14 sessions, each an hour long, and consisted primarily of attention tasks (e.g., visual scanning, visual number, and letter searches). As predicted, following treatment, accuracy and speed on the trained cognitive tasks improved, as did performance of untrained language comprehension (e.g., sentence comprehension) and production (e.g., picture description) tasks. Whether the cognitive training solely contributed to these language gains, however, cannot be established definitively: Whereas the client was most likely beyond spontaneous recovery (i.e., more than 1 year after aphasia onset), there was ample language stimulation and practice during the cognitive training sessions via natural discourse (e.g., greetings, instructions, feedback). Thus, due to the inherent presence of language in most therapy procedures, it is difficult to identify the degree of improvement, which can be attributed to a cognitive treatment. There are also special considerations when training bilingual clients use of compensatory strategies. For example, bilingual speakers with and without aphasia often successfully use one language to cue another, and these self-corrections most frequently involve producing the target word in the wrong language, intentionally or unintentionally, to cue the correct target (Goral et al., 2006; Roberts & Deslauriers, 1999). Given that explicitly teaching self-cuing is a successful word retrieval strategy among clients with aphasia in general (e.g., Abel, Schultz, Radermacher, Willmes, & Huber, 2005), utilizing these self-corrections, the powerful connection of cognates, or both could be effective strategies for bilingual speakers with aphasia, especially if they live in a bilingual community. Additionally, when these selfcorrection attempts are unsuccessful, use of a cognate, albeit in the wrong language, may still be understood, even by someone who does not speak the wrong language (Roberts & Deslauriers, 1999). Although unintended code switching can interfere with communication intent, its use could be harnessed as a compensatory strategy (Muoz et al., 1999), particularly if the client with aphasia lives in a bilingual community. In some instances, it may be most efficient to adapt to the clients strengths by training daily communication partners to understand cognates in the other language or to accept those cognates commonly used as semantic paraphasias by the client. Family members, caregivers, or even coworkers could also be taught vocabulary commonly used within the language with which they are unfamiliar but that the client uses. Viewed from this perspective, code switching can enhance functional communication. To determine if this is an appropriate option, SLPs need to identify their clients communication goals,

communication environment at home, and ability to manipulate volitionally code switching. Although augmentative and alternative communicative device use is an option for monolingual individuals with aphasia, a review of the literature indicates nominal exploration of this treatment approach for bilingual speakers with aphasia. Spanish versions of one digitized speech device, the SpringBoard Communication Aid by Prentke Romich Company, have recently been developed and used with bilingual clients (Cross, 2005). This system, designed for persons with apraxia and for children, is available in two Spanish forms, that of Spain and the U.S. domestic market, with a vocabulary set allowing for spontaneous speech. Additionally, device prompts and support materials are provided in Spanish as well as English, so the system can be switched between the two languages. Despite such device availability, SLPs should be aware of cultural barriers to such technological solutions, as ethnic culture and non-English language backgrounds are often barriers to successful device use (Reed & Newton, 2004).

Summary and Future Directions

Because individuals who speak more than one language are one of the fastest growing segments of the U.S. population and many of these individuals are at risk for developing medical conditions that can produce aphasia and other communication disorders, there is a growing need to ensure that SLPs are adequately prepared to serve the needs of the bilingual population. This review, which focused primarily on research pertaining to Spanish-English bilingual aphasia due to the dramatic increase in Spanish-English speakers in the United States and the considerable research available focusing on these languages, indicates a dire need for further empirical investigation of many issues pertaining to understanding and managing bilingual aphasia. First, debate persists regarding how best to quantify and qualify bilingualism and other linguistic concepts (e.g., cognate status) unique to bilingual speakers. It also is not yet clear how certain linguistic considerations (e.g., the nature and relationship of each languages lexicon) and cognitive factors (e.g., attention, inhibition) influence language selection and use in healthy bilingual speakers; consequently, how these factors contribute to the variety of recovery patterns observed in bilingual aphasia requires further exploration. Second, most bilingual aphasia research has involved relatively weak study designs (e.g., case descriptions, small group designs) and often failed to control for confounding variables such as differing levels of premorbid language abilities (e.g., Bates et al., 1987). Thus, replication and extension of these studies are needed. Lastly, there has been limited development and critical evaluation of assessment and treatment protocols designed to meet the unique characteristics and needs of bilingual clients with aphasia. For example, in research and clinical practice, no consistent means have yet been established to evaluate the pre- and postmorbid language abilities of bilingual clients; regular use of specific interview and/or rating tools and linguistic and cognitive tests may make comparison across future studies more meaningful and evaluation of clients more reliable. Likewise a scant

312 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

treatment literature contains mixed findings regarding how to maximize cross-linguistic generalization and has so far focused almost exclusively on remediation of lexical semantic abilities. In conclusion, demographic data indicate an urgent need for researchers and clinicians to address the many empirical questions and management needs identified in this review. ASHAs (2003) Code of Ethics requires SLPs to evaluate the effectiveness of their services and only provide services when benefit can reasonably be expected. With nominal research currently available regarding managing bilingual aphasia, SLPs are left to make clinical decisions without the appropriate support. Additional empirical investigation, therefore, is crucial to ensure ethical provision of services to this rapidly growing client population.

References
Abel, S., Shultz, A., Radermacher, I., Willmes, K., & Huber, W. (2005). Decreasing and increasing cues in naming therapy for aphasia. Aphasiology, 19, 831848. Aglioti, S., Beltramello, A., Girardi, F., & Fabbro, F. (1996). Neurolinguistic and follow-up study of an unusual pattern of recovery from bilingual subcortical aphasia. Brain, 119, 15511564. American Heart Association, Statistics Committee and Stroke Statistics Subcommittee. (2005). Heart Disease and Stroke Statistics2006 Update. Circulation: Journal of the American Heart Association. Retrieved January 27, 2006, from [Link] [Link] /cgi/content /short /113/6/e85. American Speech-Language-Hearing Association. (1989). Bilingual speech-language pathologists and audiologists: Definition [Relevant paper]. Available from [Link]/policy. American Speech-Language-Hearing Association. (1991, September). Cultural diversity in the elderly population. Asha, 33(9), 66. American Speech-Language-Hearing Association. (2003). Code of ethics (revised). Available from [Link]/policy. Ardila, A. (1991). Errors resembling semantic paralexias in Spanish-speaking aphasics. Brain and Language, 41, 437445. Ardila, A. (1998). Clinical forum: Bilingualism: A neglected and chaotic area. Aphasiology, 12, 131134. Ardila, A. (2001). The manifestation of aphasic symptoms in Spanish. Journal of Neurolinguistics, 14, 337347. Ardila, A., Rosselli, M., & Ostrosky-Solis, F. (1996). Agraphia in the Spanish language. Aphasiology, 10, 723739. Ardila, A., Rosselli, M., & Pinzon, O. (1989). Alexia and agraphia in Spanish speakers. In A. Ardila & F. Ostrosky-Solis (Eds.), Brain organization of language and cognitive processes (pp. 147175). New York: Plenum Press. Baker, R. (1993). The assessment of language impairment in elderly bilingual speakers and second language speakers in Australia. Language Testing, 10, 25276. Basso, A., & Paulin, M. (2003). Semantic errors in naming, repetition, spelling and drawing from memory: A new Italian case. Neurocase, 9, 109117. Bates, E., Friederici, A., & Wulfeck, B. (1987). Grammatical morphology in aphasia: Evidence from three languages. Cortex, 23, 545574. Benton, A. L., & Hamsher, K. D. (1994). Examen de Afasia Multilingue [Multilingual Aphasia Exam]. San Antonio, TX: The Psychological Corporation. Boller, F., Verny, M., Hugonot-Diener, L., & Saxton, J. (2002). Clinical features and assessment of severe dementia. A review. European Journal of Neurology, 9, 125136.

Bub, D., & Kertesz, A. (1982). Deep agraphia. Brain and Language, 17, 146165. Caramazza, A., & Brones, I. (1980). Semantic classification by bilinguals. Canadian Journal of Psychology, 34, 7781. Chan, R. C., Hoosain, R., & Lee, T. M. C. (2002). Reliability and validity of the Cantonese version of the test of everyday attention among normal Hong Kong Chinese: A preliminary report. Clinical Rehabilitation, 16, 900909. Chan, R. C., & Manly, T. (2002). The application of dysexecutive syndrome measures across cultures: Performance and checklist assessment in neurologically healthy and traumatically brain-injured Hong Kong Chinese volunteers. Journal of the International Neuropsychological Society, 8, 771780. Chen, H.-C., & Leung, Y.-S. (1989). Patterns of lexical processing in a non-native language. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 316325. Chen, H.-C., & Ng, M.-L. (1989). Semantic facilitation and translation priming effects in Chinese-English bilinguals. Memory and Cognition, 17, 454462. Costa, A., & Santesteban, M. (2004). Lexical access in bilingual speech production: Evidence from language switching in highly proficient bilingual speakers and L2 learners. Journal of Memory and Language, 50, 491511. Costa, A., Santesteban, M., & Cao, A. (2005). On the facilitatory effects of cognate words in bilingual speech production. Brain and Language, 94(1), 94103. Cristoffanini, P., Kirsner, K., & Milech, D. (1986). Bilingual lexical representation: The status of Spanish-English cognates. The Quarterly Journal of Experimental Psychology, 38A, 367393. Cross, R. (2005). Spanish language voice output application program using digitized speech. Paper presented at the Center on Disabilities Technology and Persons with Disabilities Conference, California State University. Retrieved January 28, 2006, from [Link]/cod/conf/2005/proceedings/[Link]. Cuetos, F., & Mitchell, D. C. (1988). Cross-linguistic differences in parsing: Restrictions on the use of the Late Closure strategy in Spanish. Cognition, 30, 73105. De Bleser, R., Dupont, P., Bormans, G., Speelman, D., Mortelmans, L., & Debrock, M. (2003). The organisation of the bilingual lexicon: A PET study. Journal of Neurolinguistics, 16, 439456. de Diego Balaguer, R., Costa, A., Sebastin-Galles, N., Juncadella, M., & Caramazza, A. (2004). Regular and irregular morphology and its relationship with agrammatism: Evidence from two Spanish-Catalan bilinguals. Brain and Language, 91, 212222. de Groot, A. M. B., & Nas, G. L. J. (1991). Lexical representation of cognates and noncognates in compound bilinguals. Journal of Memory and Language, 30, 90123. Drouesn, J., & Beauvois, M. F. (1979). Phonological processing in reading: Data from alexia. Journal of Neurology, Neurosurgery, and Psychiatry, 42, 11251132. Dijkstra, T., & van Heuven, W. J. (1998). The BIA model and bilingual word recognition. In J. Grainger & A. M. Jacobs (Eds.), Localist connectionist approaches to human cognition (pp. 189226). Mahwah, NJ: Erlbaum. Dufour, R., & Kroll, J. F. (1995). Matching words to concepts in two languages: A test of the concept mediation model of bilingual representation. Memory & Cognition, 23(2), 166180. Edmonds, L., & Kiran, S. (2004). Confrontation naming and semantic relatedness judgements in Spanish/English bilingual speakers. Aphasiology, 18, 567579. Edmonds, L., & Kiran, S. (2006). Effect of semantic naming treatment on crosslinguistic generalization in bilingual aphasia.
Lorenzen & Murray: Bilingual Aphasia Review

313

Journal of Speech, Language, and Hearing Research, 49, 729748. Fabbro, F. (1995). The neurolinguistics of bilingualism: An introduction. Hove, England: Psychology Press. Fabbro, F. (2001a). The bilingual brain: Bilingual aphasia. Brain and Language, 79, 201210. Fabbro, F. (2001b). The bilingual brain: Cerebral representation of languages. Brain and Language, 79, 21122. Fabbro, F., & Paradis, M. (1995). Differential impairments in four multilingual clients with subcortical lesions. In M. Paradis (Ed.), Aspects of bilingual aphasia (pp. 139176). New York: Elsevier. Fabbro, F., Skrap, M., & Aglioti, S. (2000). Pathological switching between languages after frontal lesions in a bilingual client. Journal of Neurology, Neurosurgery, and Psychiatry, 68, 650652. Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Minimental state: A practical method for grading the cognitive state of clients for the clinician. Journal of Psychiatric Research, 12, 189198. Galvez, A., & Hinckley, J. (2003). Transfer patterns of naming treatment in a case of bilingual aphasia. Brain and Language, 87(1), 173174. Garca-Albea, J. E., & Sanchez-Bernardos, M. L. (1986). Adaptation into Spanish of H. Goodglass & E. Kaplan, Evaluacin de la afasia y trastornos relacionados [The assessment of aphasia and related disorders]. Madrid, Spain: Editorial Mdica Panamericana. Garca-Caballero, A., Garca-Lado, I., Gonzlez-Hermida, J., Recimil, M., Area, R., Manes, F., et al. (2006). Validation of the Spanish version of Addenbrookes Cognitive Examination in a rural community in Spain. International Journal of Geriatric Psychiatry, 21, 239245. Garca-Caballero, A., Recimil, M. J., Garca-Lado, I., Gayoso, P., Cadarso-Surez, C., Gonzlez-Hermida, J., et al. (2006). ACE clock scoring: A comparison with eight standard correction methods in a population of low educational level. Journal of Geriatric Psychiatry and Neurology, 19, 216219. Goodglass, H., & Kaplan, E. (1986). La evaluacin de la afasia y de trastornos relacionados [The assessment of aphasia and related disorders] (2nd ed.). Madrid, Spain: Editorial Medica Panamericana. Goodglass, H., Kaplan, E., & Barresi, B. (2001). Boston Diagnostic Aphasia Examination (3rd ed.). New York: Lippincott, Williams, & Wilkins. Goral, M., Levy, E. S., Obler, L. K., & Cohen, E. (2006). Crosslanguage lexical connections in the mental lexicon: Evidence from a case of trilingual aphasia. Brain and Language, 98, 235247. Green, D. W. (1986). Control, activation, and resource: A framework and a model for the control of speech in bilingual speakers. Brain and Language, 27, 210233. Green, D. W. (2005). The neurocognition of recovery patterns in bilingual speakers with aphasias. In J. F. Kroll & M. B. de Groot (Eds.), Handbook of bilingualism: Psycholinguistic perspectives (pp. 516530). New York: Oxford University Press. Grodzinsky, Y. (2000). The neurology of syntax: Language use without Brocas area. Behavioral and Brain Sciences, 23, 171. Grosjean, F. (1985). Polyglot aphasics and language mixing: A comment on Perecman (1984). Brain and Language, 26, 349355. Grosjean, F. (1989). Neurolinguists, beware! The bilingual is not two monolingual speakers in one person. Brain and Language, 36(1), 315. Grosjean, F. (1998). Studying bilinguals: Methodological and conceptual issues. Bilingualism: Language and Cognition, 1(2), 131149. Hebb, D. O. (1949). The organization of behaviour: A neuropsychological theory. New York: Wiley.

Helm-Estabrooks, N. (2001). Cognitive Linguistic Quick Test. San Antonio, TX: The Psychological Corporation. Hermans, D., Bongaerts, T., de Bot, K., & Schreuder, R. (1998). Producing words in a foreign language: Can speakers prevent interference from their first language? Bilingualism, 1(3), 213229. Hernandez, A., Dapretto, M., Mazziotta, J., & Bookheimer, S. (2001). Language switching and representation in SpanishEnglish bilingual speakers: An fMRI study. NeuroImage, 14, 510520. Hernandez, A., Martinez, A., & Kohnert, K. (2000). In search of the language switch: An fMRI study of picture naming in Spanish-English bilingual speakers. Brain and Language, 73, 421431. Holland, A. L., & Fridriksson, J. (2001). Aphasia management during the early phases of recovery following stroke. American Journal of Speech-Language Pathology, 10, 1928. Huber, W., Poeck, K., Weniger, D., & Willmes, K. (1983). Aachen Aphasia Test (AAT). Gttingen, Germany: Hogrefe. Indiana Commission on Hispanic/Latino Affairs. (2005). Report of findings & recommendations: 2004-2005. Retrieved August 28, 2006, from [Link]/ichla /reports/[Link]. Iribarren, C. (2007). Description and detection of acquired dyslexia and dysgraphia in Spanish. In J. G. Centeno, R. T. Anderson, & L. K. Obler (Eds.), Communication disorders in Spanish speakers (pp. 231242). Toronto, Ontario, Canada: Multilingual Matters. Iribarren, C., Jarema, G., & Lecours, R. (1999). Lexical reading in Spanish: Two cases of phonological dyslexia. Applied Psycholinguistics, 20, 407428. Kaplan, E. F., Goodglass, H., & Weintraub, S. (1986). Test de Vocabulario de Boston [Boston Naming Test]. Madrid, Spain: Panamericana. Kay, J., Lesser, R., & Coltheart, M. (1992). PALPA: Psycholinguistic Assessment of Language Processing in Aphasia. Hove, England: Erlbaum. Kertesz, A., Pascual-Leone, P., & Pascual-Leone, G. (1990). Western Aphasia Battery en versin y adaptacin castellana [ Western Aphasia BatterySpanish Version]. Valencia: Nau Libres. Kim, K. H. S., Relkin, N. R., Lee, K.-M., & Hirsch, J. (1997). Distinct cortical area associated with native and second languages. Nature, 388, 171174. Kiran, S., & Edmonds, L. A. (2004). Effect of semantic naming treatment on crosslinguistic generalization in bilingual aphasia. Brain and Language, 91(1), 7577. Kiran, S., & Tuchtenhagen, J. (2005). Imageability effects in normal SpanishEnglish bilingual adults and in aphasia: Evidence from naming to definition and semantic priming tasks. Aphasiology, 19, 315327. Kirsner, K., Lalor, E., & Hird, K. (1993). The bilingual lexicon: Exercise, meaning and morphology. In R. Schreuder & B. Weltens (Eds.), The bilingual lexicon (pp. 215246). Amsterdam: John Benjamins. Kohnert, K. (2004). Cognitive and cognate-based treatments for bilingual aphasia: A case study. Brain and Language, 91, 294302. Kohnert, K., Hernandez, A. E., & Bates, E. (1998). Bilingual performance on the Boston Naming Test: Preliminary norms in Spanish and English. Brain and Language, 65, 422440. Kroll, J. F., & Curley, J. (1988). Lexical memory in novice bilinguals: The role of concepts in retrieving second language words. In M. Gruneberg, P. Morris, & R. Sykes (Eds.), Practical aspects of memory (Vol. 2, pp. 389395). London: Wiley. Kroll, J. F., & de Groot, M. B. (1997). Lexical and conceptual memory in the bilingual: Mapping form to meaning in two languages. In M. B. de Groot & J. F. Kroll (Eds.), Tutorials in

314 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

bilingualism: Psycholinguistic perspectives (pp. 169199). Mahwah, NJ: Erlbaum. Kroll, J. F., & Stewart, E. (1990, November). Concept mediation in bilingual translation. Paper presented at the 31st Annual Meeting of the Psychonomic Society, New Orleans. Kroll, J. F., & Stewart, E. (1994). Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. Journal of Memory and Language, 33, 149174. Laganaro, M., & Overton Venet, M. (2001). Acquired alexia in multilingual aphasia and computer-assisted treatment in both languages: Issues of generalization and transfer. Folia Phoniatrica et Logopaedica, 53, 135144. Lalor, E., & Kirsner, K. (2001). The role of cognates in bilingual aphasia: Implications for assessment and treatment. Aphasiology, 15, 10471056. Levelt, W. J. M. (1999). Models of word production. Trends in Cognitive Sciences, 3, 223232. MacIntyre, P. D., Noels, K. A., & Clement, R. (1997). Biases in self-ratings of second language proficiency: The role of language anxiety. Language Learning, 47, 265287. MacWhinney, B., Bates, E., & Kiegl, R. (1984). Cue validity and sentence interpretation in English, German, and Italian. Journal of Verbal Learning and Verbal Behavior, 23, 127150. Marian, V., Blumenfeld, H. K., & Kaushanskaya, M. (2007). The language experience and proficiency questionnaire ( LEAP-Q): Assessing language profiles in bilinguals and multilinguals. Journal of Speech, Language, and Hearing Research, 50, 941967. Marin, P., Abutalebi, J., Engelborghs, S., & De Deyn, P. P. (2005). Pathophysiology of language switching and mixing in an early bilingual child with subcortical aphasia. Neurocase, 11, 385398. Marshall, J. C., & Newcombe, F. (1973). Patterns of paralexia: A psycholinguistic approach. Journal of Psycholinguistics, 2, 175199. Mathuranath, P. S., Nestor, P. J., Berrios, G. E., Rakowicz, W., & Hodges, J. R. (2000). A brief cognitive test battery to differentiate Alzheimers disease and frontotemporal dementia. Neurology, 55, 16131620. Mcnamara, J. (1969). How can one measure the extent of a persons bilingual proficiency? In L. Kelly (Ed.), Description and measurement of bilingualism (pp. 8097). Toronto, Ontario, Canada: University of Toronto Press. Menn, L., OConnor, M., Obler, L., & Holland, A. (1995). Non-fluent aphasia in a multilingual world. Philadelphia: John Benjamins. Minkowski, M. (1983). A clinical contribution to the study of polyglot aphasia especially with respect to Swiss-German. In M. Paradis (Ed.), Readings on aphasia in bilinguals and polyglots (pp. 205232). Montreal, Quebec, Canada: Didier. (Original work published 1927) Morgenstern, L., Smith, M., Lisabeth, L. D., Risser, J. M., Uchino, K., Garcia, N., et al. (2004). Excess stroke in Mexican Americans compared with non-Hispanic Whites: The Brain Attack Surveillance in Corpus Christi Project. American Journal of Epidemiology, 160, 376383. Muoz, M. L., & Marquardt, T. P. (2003). Picture naming and identification in bilingual speakers of Spanish and English with and without aphasia. Aphasiology, 17, 11151132. Muoz, M. L., Marquardt, T., & Copeland, G. (1999). A comparison of the code-switching patterns of speakers with aphasia and neurologically normal bilingual speakers of English and Spanish. Brain and Language, 66, 249274. Murray, L. L. (1999). Attention and aphasia: Theory, research and clinical implications. Aphasiology, 13, 91111.

Napps, S. E., & Fowler, C. A. (1987). Formal relationships among words and the organization of the mental lexicon. Journal of Psycholinguistic Research, 16, 257272. Ostrosky-Solis, F., Marcos-Ortega, J., Ardila, A., Roselli, M., & Palacios, S. (1999). Syntactic comprehension in Brocas speakers with aphasia Spanish-speakers: Null effects of word order. Aphasiology, 13, 553571. Paivio, A., Clark, J. M., & Lambert, W. (1988). Bilingual dualcoding theory and semantic repetition effects on recall. Journal of Experimental Psychology, 14(1), 163172. Paivio, A., & Desrochers, A. (1980). A dual-coding approach to bilingual memory. Canadian Journal of Psychology, 33, 1728. Paradis, M. (1977). Bilingualism and aphasia. In H. Whitaker & H. Whitaker (Eds.), Studies in neurolinguistics (Vol. 3, pp. 65121). New York: Academic Press. Paradis, M. (1984). Aphasie et traduction [Aphasia and translation]. Meta Translators Journal, 24, 5767. Paradis, M. (1988). Recent developments in the study of agrammatism: Their import for the assessment of bilingual aphasia. Journal of Neurolinguistics, 3(2), 127160. Paradis, M. (2001). Bilingual and polyglot aphasia. In R. S. Berndt (Ed.), Language and aphasia (pp. 6991). Amsterdam: Elsevier Science. Paradis, M. (2004). A neurolinguistic theory of bilingualism. Amsterdam: John Benjamins. Paradis, M., Goldblum, M., & Abidi, R. (1982). Alternate antagonism with paradoxical translation behaviour in two bilingual speakers with aphasia clients. Brain and Language, 15, 5569. Paradis, M., & Libben, G. (1987). The assessment of bilingual aphasia. Hillsdale, NJ: Erlbaum. Perani, D., Paulesu, E., Galles, N. S., Dupoux, E., Dehaene, S., Bettindardi, V., et al. (1998). The bilingual brain: Proficiency and age of acquisition of the second language. Brain, 121, 18411852. Perecman, E. (1984). Spontaneous translation and language mixing in a polyglot aphasic. Brain and Language, 23, 4363. Pontn, M. O., Satz, P., Herrera, L., Ortiz, F., Urrutia, C. P., Young, R., et al. (1996). Normative data stratified by age and education for the Neuropsychological Screening Battery for Hispanics (NeSBHIS): Initial Report. Journal of the International Neuropsychological Society, 2, 96104. Potter, M., So, K., Von Eckardt, B., & Feldman, L. (1984). Lexical and conceptual representation in beginning and proficient bilinguals. Journal of Verbal Learning and Verbal Behavior, 23, 2338. Price, C. J., Green, D. W., & von Studnitz, R. (1999). A functional imaging study of translation and language switching. Brain, 122, 22212235. Randolph, C. (1998). Repeatable Battery for the Assessment of Neuropsychological Status. San Antonio, TX: The Psychological Corporation. Reed, M., & Newton, M. (2004, March). Examining the effectiveness of assistive technology and related services for ethnic minorities. Paper presented at the Center on Disabilities Technology and Persons With Disabilities Conference. Retrieved January 28, 2006, from [Link]/cod/conf/2004/proceedings/ [Link]. Rey, G. J., Feldman, E., Hernandez, D., Levin, B., RivasVazquez, R., Nedd, K., & Benton, A. L. (2001). Application of the multilingual aphasia examination-Spanish in the evaluation of Hispanic clients post closed-head trauma. The Clinical Neuropsychologist, 15(1), 1318. Rey, G. J., Sivan, A. B., & Benton, A. L. (1991). Multilingual Aphasia ExaminationSpanish Version. Lutz, FL: Psychological Assessment Resources.
Lorenzen & Murray: Bilingual Aphasia Review

315

Reyes de Beaman, S., Beaman, P. E., Garcia-Pea, C., Villa, M. A., Heres, J., Crdoa, A., & Jagger, C. (2004). Validation of a modified version of the Mini-Mental State Examination (MMSE) in Spanish. Aging Neuropsychology and Cognition, 11, 111. Roberts, P. M. (1998). Clinical research needs and issues in bilingual aphasia. Aphasiology, 12, 119146. Roberts, P. M. (2001). Aphasia assessment and treatment for bilingual and culturally diverse clients. In R. Chapey (Ed.), Language intervention strategies in aphasia and related neurogenic communication disorders (pp. 208232). New York: Lippincott, Williams & Wilkins. Roberts, P. M., de la Riva, J., & Rhaume, A. (1997, May). Effets de lintervention dans une langue pour lanomie bilinge [Effects of language intervention for bilingual anomia /aphasia]. Paper presented at the annual Canadian Association of SpeechLanguage Pathology and Audiology conference, Toronto, Ontario, Canada. Roberts, P. M., & Deslauriers, L. (1999). Picture naming of cognate and non-cognate nouns in bilingual aphasia. Journal of Communication Disorders, 32(1), 122. Robertson, I. H., Ward, T., Ridgeway, V., & Nimmo-Smith, I. (1994). The Test of Everyday Attention. Bury St. Edmunds, England: Thames Valley Test. Ross, S. (1998). Self-assessment in second language testing: A meta-analysis and analysis of experiential factors. Language Testing, 15(1), 120. Rosselli, M., Ardila, A., Florez, A., & Castro, C. (1990). Normative data on the Boston Diagnostic Aphasia Examination in a Spanish-Speaking population. Journal of Clinical and Experimental Neuropsychology, 12, 313322. Sampson, G. (1985). Writing systems. Stanford, CA: Stanford University Press. Saxton, J., McGonigle-Gibson, K., Swihart, A., Miller, M., & Boller, F. (1990). Assessment of the severely impaired client: Description and validation of a new neuropsychological test battery. Psychological Assessment: Journal of Consulting Clinical Psychology, 2, 298303. Segalowitz, S. J., Segalowitz, N. S., & Wood, A. G. (1998). Assessing the development of automaticity in second language word recognition. Applied Psycholinguistics, 19, 5367. Semel, E., Wiig, E. H., & Secord, W. A. (1997). Clinical Evaluation of Language Fundamentals, Third Edition (Spanish Version). San Antonio, TX: The Psychological Corporation. Sparks, R., Helm, N., & Albert, M. (1974). Aphasia rehabilitation resulting from Melodic Intonation Therapy. Cortex, 10, 303316. Ullman, M. (2001). A neurocognitive perspective on language: The declarative/procedural model. Nature Reviews, 2, 717726. U.S. Bureau of the Census. (2003). Language use and Englishspeaking ability: 2000. U.S. Census Bureau American Fact Finder. Retrieved January 25, 2006, from [Link] [Link]/servlet /QTTable?_ bm=y&-geo_ id=01000US&qr_ name=DEC_ 2000_ SF3_U_QTP17&-ds_ name=DEC_ 2000_ SF3_U.

U.S. Department of Health and Human Services, Agency for Healthcare Research and Quality. (2004). National Healthcare Disparities Report. Retrieved August 24, 2006, from [Link]/qual/nhdr04/nhdr04htm. Velasquez, M., Arcos-Burgos, M., Toro, M. E., Castao, A., Madrigal, L., Moreno, S., et al. (2000). Analisis factorial y discriminante de variables neuropsicologicas en la demencia tipo Alzheimer de inicio tardio, familiar y esporadica [Analysis of discriminating factor and neuropsychological variables in Alzheimer-type dementia: Late onset, familial and sporadic]. Revista de Neurologia, 31, 501506. Wai-kwong Man, D., & Li, R. (2001). Assessing Chinese adults memory abilities: Validation of the Chinese version of the Rivermead Behavioral Memory Test. Clinical Gerontologist, 24(3/4), 2736. Watamori, T., & Sasnuma, S. (1976). The recovery process of a bilingual speakers with aphasia. Journal of Communication Disorders, 9, 157166. Watamori, T., & Sasnuma, S. (1978). The recovery process of two English-Japanese bilingual speakers with aphasias. Brain and Language, 6, 127140. Wechsler, D. (1984). Test de Inteligencia para Adultos WAIS Manual. Adaptacin de Buenos Aires [Wechsler Adult Intelligence Scale (WAIS). Buenos Aires Version]. Buenos Aires, Argentina: Paidos. Wilson, B. A., Alderman, N., Burgess, P., Emslie, H., & Evans, J. (1996). Behavioral Assessment of the Dysexecutive Syndrome. Bury St. Edmunds, England: Thames Valley Test. Wilson, B. A., Cockburn, J., & Baddeley, A. (1985). The Rivermead Behavioral Memory Test. Bury St. Edmunds, England: Thames Valley Test. Woodcock, R. W., & Muoz-Sandoval, A. F. (1996a). Batera Woodcock-Muoz: Pruebas de aprovechamiento-Revisada [Woodcock-Muoz Tests of Achievement. Revised]. Itasca, IL: Riverside. Woodcock, R. W., & Muoz-Sandoval, A. F. (1996b). Batera Woodcock-Muoz: Pruebas de habilidad cognitiva-Revisada [Woodcock-Muoz Tests of Cognitive Ability Revised]. Itasca, IL: Riverside. Received March 3, 2007 Accepted December 18, 2007 DOI: 10.1044/1058-0360(2008/026) Contact author: Bonnie Lorenzen, Department of Speech and Hearing Sciences, 200 South Jordan Avenue, Bloomington, IN 47405-7002. E-mail: blorenze@[Link].

316 American Journal of Speech-Language Pathology Vol. 17 299317 August 2008

Appendix Examples of Cognitive and Language Tests Available in Other Languages

Test Aachen Aphasia Test (Huber et al., 1983) Addenbrookes Cognitive Examination (Mathuranath, Nestor, Berrios, Rakowicz, & Hodges, 2000) Behavioral Assessment of the Dysexecutive Syndrome (Wilson, Alderman, Burgess, Emslie, & Evans, 1996) Boston Diagnostic Aphasia Exam (Goodglass & Kaplan, 1986) Boston Naming Test (Kaplan, Goodglass, & Weintraub, 1986) Clinical Evaluation of Language Fundamentals (Semel, Wiig, & Secord, 1997) Cognitive Linguistic Quick Test (Helm-Estabrooks, 2001) Consortium to Establish a Registry for Alzheimer s Disease (Velasquez et al., 2000) Mini-Mental State Examination (Folstein, Folstein, & McHugh, 1975) Neuropsychological Screening Battery for Hispanics (Pontn et al., 1996) Psycholinguistic Assessment of Language Processing in Aphasia (Kay et al., 1992) Repeatable Battery for the Assessment of Neuropsychological Status (Randolph, 1998) Rivermead Behavioral Memory Test (Wilson, Cockburn, & Baddeley, 1985) Severe Impairment Battery (Saxton, McGonigle-Gibson, Swihart, Miller, & Boller, 1990) Test of Everyday Attention (Robertson, Ward, Ridgeway, & Nimmo-Smith, 1994) Wechsler Adult Intelligence Scale (Wechsler, 1984) Western Aphasia BatterySpanish (Kertez et al., 1990). Woodcock-Muoz: Pruebas de aprovechamientoRevisada (Woodcock & Muoz-Sandoval, 1996a) Woodcock-Muoz: Pruebas de habilidad cognitivaRevisada (Woodcock & Muoz-Sandoval, 1996b) Description, translations, and/or normative data Multiple languages Adapted clock drawing tests in Spanish (Garca-Caballero, Garca-Lado, et al., 2006), as well as other languages (see Garca-Caballero, Recimil, et al., 2006) Chinese version (Chan & Manly, 2002) Spanish (Garca-Albea, & Sanchez-Bernardos, 1986), and other language versions Spanish version with normative data (Kohnert, Hernandez, & Bates, 1998) Spanish version Spanish version with normative data Spanish cognitive test battery Spanish version with normative data (Reyes de Beaman et al., 2004), and additional language versions Spanish cognitive test battery Spanish version Spanish version Chinese version (Wai-kwong Man & Li, 2001) Available in several languages such as French, Italian, and Swedish (see Boller, Verny, Hugonot-Diener, & Saxton, 2002) Cantonese version (Chan, Hoosain, & Lee, 2002) Spanish version and other languages Spanish version Spanish test of academic skills for all ages through geriatric Spanish cognitive test battery

Lorenzen & Murray: Bilingual Aphasia Review

317