The definition of bilingualism has been long debated (Lorenzon and Murray, 2008). However, although it has been claimed by previous researchers that a requirement of bilingualism is to have equal abilities across both languages, there is now consensus in the literature that it is not necessary to have complete proficiency in both languages in order to be bilingual (Wei, 2007). This is due to the fact that individuals tend to use different languages for different areas of life, e.g. work vs home life, and, therefore, across these different areas, levels of proficiency will differ (Lorenzon and Murray, 2008). Approximately 60% of the global population are bi- or multiilingual (Kiran, Sandberg, Gray, Ascenso and Kester, 2013) and Kotik-Friedgut (2001), suggests that bilinguals can be separated into two groups when taking account of age of acquisition: successive bilinguals and simultaneous bilinguals. Successive bilinguals learn L1 and L2 at different stages, while simultaneous bilinguals learn their L2 (second language) and L1 (first language) at the same time during childhood. Bilinguals may experience ‘first language attrition’ which denotes a non-pathological reduction in the competency of L1, typically when an individual has moved to a predominantly L2 environment from a predominantly L1 environment and remains in an environment where L1 usage is low (Seton and Schmid, 2016).
Aphasia is defined as an acquired disorder of communication as a result of brain damage (Chapey, 2008) and is described by Ansaldo and Ghazi-Saidi (2014) as an inability to interpret, formulate or retrieve the symbols which make up language. It was put forward by Paradis (2004), that we should consider aphasia as a disorder which restricts parts of a language system instead of thinking that the language system itself has been damaged by aphasia. Aphasia constitutes an impairment of linguistic capability, thus pragmatics and metalinguistic knowledge would not be affected as they are stored in separate areas of the brain. Two main groups of aphasia exist which are categorized depending on the site of brain lesion(s): non-fluent and fluent. In non-fluent aphasia, expressive language abilities are primarily affected, whereas in fluent aphasia, the main area of difficulty is language comprehension (Altarriba & Heredia, 2018). The most common cause of aphasia is stroke (Jordan & Kaiser, 1996) and approximately 20% of stroke survivors go on to develop aphasia (Dobkin, 2005). As there has been an increase in the number of stroke survivors in recent years, the number of bilingual adults who have acquired aphasia has grown, which presents a challenge to SLT services (Koumanidi Knoph, 2011). Aphasia will affect bilinguals in different ways and may only affect one language, which is known as selective aphasia, or both languages. It has been suggested in a number of studies that selective aphasia in one language may be due to issues with language control (e.g. Green, Grogan, Crinion, Ali, Sutton & Price, 2010). Nevertheless, it was found by Fabbro (2001) that the type of aphasia, e.g. expressive, will remain the same when both languages have been affected, although the presentation of aphasia may be different because of differences in typology.
Fabbro (2001), outlines the three phases of aphasia which are important to recognize when assessing and treating patients. Symptoms can alter without any intervention while the patients is going through the phases: 1) The acute phase: lasts around 4-weeks post onset, 2) The lesion phase: lasts for up to 6 months and spontaneous recovery may occur in this stage, 3) The chronic/late phase: this starts several months after onset and persists throughout the lifetime of the patient.
Recovery patterns in patients with bilingual aphasia (PWBA)
In terms of recovery, it has been found that PWBA may exhibit either parallel or non-parallel recovery. When parallel recovery occurs, L1 and L2 will recover to the same degree in relation to the individual’s competency before the onset of aphasia. However, if non-parallel recovery occurs, competency in L1 and L2 will differ after the onset of aphasia and will be different from pre-morbid competency (Kohnert, 2013). Subtypes within non-parallel recovery include ‘selective recovery’ meaning that only one language has recovered (Berthier, Starkstein, Lylyk, and Leiguarda, 1990), ‘alternate antagonism’ describes a PWBA who has issues with the alternation of the two languages i.e. one language may be easier to retrieve than another on certain days and vice versa (Nilipour and Ashayeri, 1989), ‘successive recovery’ denotes the recovery of one language before the other (Paradis, 1983) and ‘unintentional language switching’ is defined as pathological code-switching between the two languages (Fabbro, Skrap, and Aglioti, 2000). Although studies have found that parallel recovery occurs more frequently in bilingual patients than non-parallel recovery (e.g. Fabbro, 2001), there has been significant interest in studies concerned with non-parallel recovery as explanations are sought as to the reason one language has been more impaired than the other (Kohnert, 2013).
There is disagreement among academics regarding how patterns of language are recovered by bilinguals with aphasia. It was put forward by Pitres (1895), that the language which is most used by the PWBA at the time of aphasia onset will recover as the components in the brain which function in this language are more intimately linked. Nonetheless, no discernable evidence exists which means that Pitre’s rule could be related to every case of bilingual aphasia (Pearce, 2005). In contrast, Ribot’s law (1881) claims that the PWBA will recover their native language. However, only bilingual individuals who are not completely fluent in languages acquired later in life have been found to conform to Ribot’s Law (Pearce, 2005). Kohnert (2013), reported the case of a bilingual Italian-English patient who showed damage in all language modalities across both languages. It was found that the patient’s speaking abilities were better in English (L2), which is consistent with Pitre’s rule, while assessment of reading skills showed that Italian (L1) produced better results, adhering to Ribot’s law. Consequently, these results show that no single rule can explain the recovery patterns of all language abilities for one individual.
Uncertainty still exists regarding how different languages are stored in the brain. ‘The Linguistic Domain Approach’ proposes that multiple languages have one representation in the brain (Paradis, 1990; Moretti, Bava, Torre, Antonello, Zorzon, Zivandinov, Cazzato, 2001), whereas ‘The Language Membership Principle’ suggests that there may be different areas for language storage in the bilingual brain (Dehaene et al, 1997; Green, 2005). Therefore, in line with ‘The Language Membership Principle’, L1 and L2 would be represented in distinct areas of the brain due to the fact that they would exhibit different values of language membership (Ibrahim, 2009). Gomez-Tortosa et al. (1995), in accordance with the ‘Language Membership Principle,’ proposed that non-parallel recovery occurs due to different languages being stored in separate areas of the brain. Nonetheless, this contradicts the findings of many research studies (e.g Costa and Sebastián-Gallés, 2014) demonstrating that different areas of the brain will overlap when bilinguals are processing language.
Issues affecting assessment and intervention of bilinguals with aphasia
The primary aim of assessing PWBA is to investigate impaired and intact communication skills across both languages (Altarriba and Heredia, 2018). It is well established in the literature that bilinguals with aphasia need to be assessed in both languages as one language may be more severely affected than another (Paradis, 1995) and pre-morbid proficiency levels need to be confirmed as a reference point to current language levels of the patient (Paradis, 2011). Issues which may affect the assessment and treatment of bilinguals with aphasia are outlined below.
Firstly, the Bilingual Aphasia Test (Paradis, 1987) is a criterion-referenced measure which provides a comparison across languages of the impairment in PWBA and is the test used most frequently to measure BA. Fabbro (1999), notes that it has been produced in 60 languages and 106 language pairs. The BAT is made up of three sections: Part A investigates the patient’s language history, as well as their family, Part B assesses features of the patient’s expressive and receptive language and Part C examines the translation abilities of the patient across languages at word and sentence level (Muñoz and Marquardt, 2008). Although the BAT is seen as an essential research tool for the assessment of PWBA who speak a variety of languages and language pairs (Nilipour and Paradis, 1995), few studies have been undertaken to analyse the BAT’s validity (Muñoz and Marquardt, 2008) and it was claimed by Roberts (1998) that we should assess the BAT’s psychometric features and its suitability for other purposes in order to make it of more clinical benefit.
Secondly, Chapey (2008), notes that at least part of the improvement that occurs in the first few weeks following the onset of aphasia is due to spontaneous recovery. Gil and Goral (2004), state that, it is unclear precisely how spontaneous recovery will affect each of the languages spoken by bilinguals with aphasia, but it is probable that it will affect both languages. As we do not currently understand the interaction between recovery which occurs due to treatment and spontaneous recovery, it is therefore extremely difficult, while treatment is occurring, to separate the two in the first few weeks after onset. This could result in the need to re-assess the patient after the spontaneous recovery period to determine if language levels have improved.
Thirdly, assessing PWBA will typically involve more than one person, e.g. an SLT and an interpreter, while a monolingual language assessment may be conducted on a one-to-one basis. Jodache, Howe and Siyambalapitiya (2015), conducted a study which focused on the experiences of participants involved in the language assessment of a Samoan–English patient with bilingual aphasia. The team consisted of two interpreters, one SLT and a family member. It was found that certain aspects resulted in the team working less effectively, including one of the interpreters conversing with the family member of the patient separately during the language assessment. Roger and Code (2011), note that these findings emphasise the importance of information sharing between SLTs and interpreters including: briefing the interpreter before and after the session, informing the interpreter regarding what to look out for in terms of errors from the patient and providing background information about communication disorders.
A practical consideration in providing bilingual aphasia therapy is which language the clinician should use while providing treatment (Gil and Goral, 2004). Historically, it was thought that treatment for PWBA ought to target only one of the languages, particularly for patients who exhibit pathological language switching (Ward, 1961; Fabbro, 2001). The rationale for targeting one language was that this would be less confusing for the patient, and so recovery time would be faster (Altarriba and Heredia, 2018). However, Kohnert (2013) suggests that only treating one language is not treating the patient in a holistic way and if giving intervention in one language causes a patient to be cut-off from their communicative environment, for instance, the benefits gained from treatment are cancelled out.
Although there is still no clear agreement in the literature over which language a clinician should target in treatment, Altarriba and Heredia (2018) outline the three options for treating bilinguals with aphasia: 1) treatment is given in the patient’s first language, 2) treatment is given in the patient’s second language, 3) treatment is given in both languages. According to Lerman, Edmonds and Goral (2018), approximately 50% of all cases of bilingual aphasia treatment that have been published have revealed that cross-linguistic transfer has occurred and it was proposed by Green (2005), that advantages unique to bilinguals should be considered when devising a therapy programme. For instance, bilinguals with aphasia who exhibit parallel recovery will often produce a correct word in their non-treated language by self-cueing in order to retrieve their chosen word (Green, 2005). Consequently, Ansaldo, Marcotte, Scherer, and Raboyeau (2008) suggest that, it may not be right to deprive a bilingual patient from using the conserved features of the language which is not being targeted. Altarriba and Heredia (2018) state that, in order to treat PWBA in a holistic way, clinicians should consider the following when providing intervention: 1) The language preference of the patient and their family as the effectiveness of the intervention may be affected by the patient’s motivation, 2) The language which is most functional for the patient e.g. it may be a requirement in their place of work to use one language more than another, 3) The available resources for intervention e.g. the possibility of using a bilingual speech and language therapist or a family member to aid in providing intervention.
Conclusion
In summary, providing bilingual aphasia assessment and therapy can be challenging for clinicians as aphasia is a complex disorder which will present differently in each patient and both languages may not recover in an identical way or to previous competency before aphasia onset. Researchers understand very little as to how certain patterns of language recovery are caused in PWBA and so it would be beneficial for more research into this area to be undertaken in order to lead clinical intervention. It is essential to adopt a patient-centred approach when working with PWBA, and not just consider the disorder from a medical standpoint, as if functional needs are not considered this could lead to the individual being excluded from certain areas of their life, leading to psychosocial consequences such as isolation.