10.1 Introduction

While there is abundant descriptive and theoretical literature on the morpho-syntactic aspects of code-switching (hereafter CS) in a variety of language pairings, the phonetic and phonological reflexes of CS remain relatively unexplored. The paucity of research on these latter properties of CS may reflect the widespread assumption that, in contrast to borrowing, CS utterances manifest an abrupt transition between the sound systems of each language. When this view is challenged, it is generally done so on the basis that it inaccurately describes the degree of phonological integration that lexical borrowings, but not code-switches, may undergo. The adaptation of loan words has received considerable attention from phonologists, but the relationship between CS and the sound system of a language has not. If borrowing and CS fall along a single continuum, as many linguists have argued, then it is possible that CS utterances, as well as borrowings, may manifest some degree of integration or convergence.

This chapter presents an overview of the extant research on the phonetics of CS and attempts to address the types of questions that a full linguistic inquiry into the phonetics/phonology of CS should explore. Much of the current literature on phonetics and CS arises from the field of psycholinguistics, where the focus is on the mechanisms underlying CS in bilinguals (see Kutas et al., Gullberg et al., this volume) rather than on understanding the role of phonetics/phonology in relation to the structural aspects of bilingual CS. By examining both the psycholinguistic and the structural aspects of the phonetics of CS, this chapter demonstrates that many of the controversies that arise in explorations of the morpho-syntax of CS exist as well for the phonetic domain. In this respect, three broad questions regarding the role of the sound system in CS can be raised:

1. (1.) Does CS have an effect on phonological/phonetic production and perception?

2. (2.) Can phonological/phonetic properties be observed to constrain CS production?

3. (3.) Is there a phonetic base or matrix language in CS?

Each of these questions has been addressed in the small body of research on the phonetics and phonology of CS, but the findings of these studies are often contradictory. Nevertheless, this chapter will advance tentative answers to these questions and address the many challenges that await future researchers in this field.

As has often been noted, there is a good bit of terminological confusion surrounding the term “code-switching.” This may be particularly true of some of the literature on CS and phonetics, where “code-switching” may refer not to the alternation of languages within a single utterance but instead to a bilingual’s performance in one language rather than the other (see Bahr and Frisch Reference Bahr, Frisch, Braun and Masthoff2002 on “code-switching” and voice identification in forensic phonetics; Hazan and Boulakia Reference Hazan and Boulakia1993 on phonetic production). The focus of this chapter will be limited to a consideration of the perception and production of bilingual speakers when they are performing simultaneously in both languages either via alternational or insertional CS (see Muysken Reference Muysken2000). The organization of this chapter is as follows: § 10.2 examines the use of phonological integration as a metric for distinguishing borrowing from CS. In § 10.3 we turn to a review and analysis of psycholinguistic “switching studies” that are largely devoted to examining bilingual perception and that rely on the notion of a phonetic base language. § 10.4 reviews the findings of a handful of recent linguistic studies on the phonetics and phonology of CS productions that, in part, advance answers and introduce new complexities into the question of whether bilinguals truly switch completely from the phonetic structure of one language to the next. The possibility that prosody constrains CS is considered in § 10.5. Finally, § 10.6 concludes with areas to be investigated and challenges for future research on CS and sound structure.

10.2 The phonology and phonetics of contact phenomena

There has been a great deal of debate within the field of contact linguistics on whether or not borrowing can be distinguished from CS on the basis of phonological structure. It is important to clarify what is intended by phonological, as opposed to phonetic, structure. Phonology is commonly held to be distinct from phonetics. Where phonological differences are envisioned as categorical, phonetic ones are seen as gradient. For instance, /b/ defines the phonemic category of a voiced stop which, depending on the language and the context, may in actuality be only partially or gradiently voiced. Similarly, L(ow)H(igh) defines the distinctive phonological category of a rising tone but the slope of the LH tone may be more or less steep depending on the distance of the interval between the pitch alignment positions of the valley and the peak. Phonological distinctions, such as /b/ or LH, are generally salient to native speakers, whereas the gradient phonetic properties of an utterance, such as more or less voicing or steeper pitch rises, are not.

In a general way, the division between phonology and phonetics is analogous to the segregation of research strands in borrowing and CS, respectively. Much of the work on borrowing is undertaken at the phonological level, analyzing broadly transcribed data to advance the notion that a borrowing conforms to the sound pattern of its recipient language. Conversely, research on the interaction between CS and sound structure invariably involves examining (or manipulating) the discrete phonetic properties of an utterance since it is assumed that code-switches should manifest only marginal cross-linguistic assimilation or, ideally, none at all. The following sections discuss, in turn, the phonology of borrowing and the phonetics of CS.

10.3 The processing of acoustic information in bilingual switching studies

Psycholinguists interested in the mechanisms that underlie bilingualism, such as lexical access, inhibitory control, and selective attention, have conducted a series of studies investigating the acoustic and phonetic properties of language switching. These works largely aim to test proposals similar to those put forth by Macnamara (Reference Macnamara1967a, Reference Macnamarab) and Macnamara and Kushnir (Reference Macnamara and Kushnir1971) that bilinguals’ control of the input (perception) operates independently from their control of the output (production). Under such a view, the input switch is said to be automatic and biased toward the language of the incoming signal. That is, speakers expect the input signal to continue in the same language, and hence their processing strategies are tuned to that language. Thus, language switching has a processing cost. On the other hand, the output switch would operate under the conscious or voluntary control of the speaker. The normal design of a switching study involves the insertion of a “guest” word into what is termed a “base” or “precursor” language that provides the language set for the input. The aim is to determine whether the base language affects the recognition, perception, or production of the guest word.

Gullberg et al. (this volume) define language switching studies as those that induce participants to switch at a predestined point in an utterance. This is distinct from CS, which is assumed be voluntary. For the purposes of this chapter, though, switching studies are additionally characterized as experiments that examine the insertion of only a single guest word into a base language utterance. Thus, from a linguistic point of view, this kind of switching may fall more toward the borrowing than the CS end of the continuum of contact phenomena. However, at least one switching study (Li Reference Li1996, discussed below) endeavors to control for these different contact phenomena by manipulating the phonological structure of the guest word.

In the phonetic realm, switching studies normally target bilingual perception and are nearly exclusively dedicated to examining one phonological parameter, the categorical perception of the voiced /b,d,g/ versus voiceless /p,t,k/ series of stop consonants. One acoustic cue for the voiced–voiceless distinction is voice onset time (VOT), which defines the interval between the burst release of the consonant and the onset of voicing of the vowel. VOT spans a continuum with different languages situating the transition between a voiced and voiceless stop at different points. In Spanish and French, voiceless stops are produced with very short VOT values and are said to be short lag stops. In English, by contrast, VOT values for voiceless stops are relatively long and such stops are produced with a period of aspiration following the consonant burst, as indicated in the waveform diagram in Figure 10.1. The gradient nature of the voicing lag makes it an ideal testing ground for perceptual switching studies since the VOT value can be manipulated either through the creation of synthetic stimuli or through edited natural speech tokens. This allows for the establishment of clear end-points; for instance, VOT values of − 60 ms would be perceived as voiced by all listeners and, at the other extreme, values of 60 ms, as voiceless. Between the two endpoints lie ambiguous stimuli that could be perceived as either voiced or voiceless. In general, phonetic switching studies have been designed to test whether the language of presentation, the base language, has an effect on the perceptual categorization of ambiguous inputs. The results have been contradictory, so it is worth considering the relevant experiments in turn.

Figure 10.1 Waveform of English cat showing long voicing lag and accompanying aspiration for initial /k/ between the vertical lines

Using synthetically generated nonce syllables, Caramazza et. al (Reference Caramazza, Yeni-Komshian, Zurif and Carbone1973), testing French–English bilinguals, and Williams (Reference Williams1977), testing Spanish–English bilinguals, found that listeners were unaffected by the language of the experimental instructions, taken as the precursor language, and that bilinguals appeared to have fixed (i.e. merged) perceptual boundaries for the voicing distinction across their two languages. Elman et al. (Reference Elman, Diehl and Buchwald1977) directly assessed the effect of the precursor language on bilingual perceptual switching, using natural stimuli embedded in either English (1a) or Spanish (1b) base language contexts, as shown by the translations equivalents in (1).

In contrast to the previous findings in VOT switching studies, Elman et al. found that bilinguals did shift their perceptual boundary in response to the precursor language. Further, the effect remained when their listeners were divided into groups reflecting different levels of bilingual proficiency. Even the highest proficiency bilinguals performed differently from the corresponding monolingual groups. The researchers hypothesized that their results differed from those of the previous studies primarily due to the use of natural versus synthetic speech tokens.

A number of subsequent studies confirm the dominance of the base language on the perception of the guest language in CS but acknowledge that there might be numerous factors – structural, contextual, and psychological – that impinge on a listener’s access of a CS word (Soares and Grosjean Reference Soares and François1984; Grosjean and Soares Reference Grosjean, Soares and Vaid1986; Grosjean Reference Grosjean1988). With respect to structural factors, Bürki-Cohen et al. (Reference Bürki-Cohen, Grosjean and Miller1989) hypothesize that the phonetic structure of the stimuli itself may have a bearing on bilingual perception during CS. They constructed two different sets of stimuli, one in which the switched tokens could be homophonous across languages (French dé “dice” and English day), and one in which the phonology provides a distinctive cue to the guest language (French ré [ʁe] and English ray [ɹe]). They edited the tokens by splicing French and English productions together to create ambiguous or hybrid stimuli for the perception tasks. As in the Elman et al. study, the stimuli were embedded into base language carrier phrases as in (2).

They found that the base language had no effect on the listener’s categorization of the language-neutral series of stimuli. The ambiguous stimuli of this series were identified as the same regardless of the precursor. However, they found a polarizing effect of the base language on the perception of the language-selective tokens. Here, the hybrid tokens were categorized more toward the guest language, in contrast to the base language. This implies that any effect of the base language is not necessarily assimilatory.

The use of the distinctive phonetic and phonotactic structure of a guest word as a perceptual cue to a language switch is also investigated by Li (Reference Li1996). Li uses phonological criteria to distinguish English borrowings from code-switches in a Chinese–English context. For instance, the English word flight is pronounced [faɪ] as a borrowing but as a CS, it retains the English phonetic and phonotatic structure [flaɪt]. Li shows that the structurally distinctive properties of a CS allow listeners to recognize an English word in a Chinese base language as quickly as monolingual English listeners do. The recognition of borrowings that are phonologically integrated into Chinese was found to take much more time. Li uses this evidence as an argument against an automatic language input switch since the precursor language does not affect the perception of a CS. His results can be seen to affirm those of Bürki-Cohen et al. (Reference Bürki-Cohen, Grosjean and Miller1989) in that a significant phonological dissimilarity between languages can apparently facilitate the recognition or perception of CS.

The studies by Bürki-Cohen et al. (Reference Bürki-Cohen, Grosjean and Miller1989) and Li (Reference Li1996) show that the effect of the precursor language on the perception of the guest language is probably not independent of the phonological properties under examination. This may be true of the acoustic level as well. Hazan and Boulakia (Reference Hazan and Boulakia1993) examine an additional phonetic cue to the voicing distinction in stops, the frequency of the first formant (F₁) at the onset of the voicing of the vowel. F₁ onset frequency can present a strong perceptual cue for the voicing distinction in English but not in French. In contrast to the VOT continuum that serves as a distinctive voicing cue in both English and French, the cue weighting of F₁ onset frequency, then, is categorically different across these languages. In their study, edited tokens of /bɛn/ and /pɛn/, real words in both English and French, were edited to have an identical VOT range but to vary in F₁ frequency at the onset of the syllable rhyme. As in previous studies, the test materials were constructed in a base language + guest word series to test the effect of switching, as in (3), and presented to French–English bilinguals who differed in language dominance.

Their results showed only a small effect of the precursor language on phoneme categorization and, for a majority of the bilinguals, the precursor language failed to affect cue-weighting at all. They tentatively conclude that language dominance, defined as the language learned first, determines cue-weighting in bilinguals.

Taken together, the results of the perceptual studies offer only tentative evidence that the base or precursor language affects the perception or recognition of the guest word. Soares and Grosjean (Reference Soares and François1984) enumerate various linguistic and psycho-social factors of CS that might impinge upon bilingual listeners’ performance in these tasks, few of which are ever taken into account in psycholinguistic studies of CS. Nevertheless, given the available evidence, it is unlikely that the base language functions as the phonetic equivalent of the morpho-syntactic matrix language (see Myers-Scotton and Jake, this volume), providing an acoustic frame for the perception of a mixed language utterance.

10.4 Laboratory research on the phonetics of CS

There are a number of conceptual issues underlying switching studies that limit their possible extension to understanding the phonetics of naturalistic CS. First, the guest language is represented only by a single syllable or word, a structure that is representative of a lexical insertion rather than an intra-sentential CS. As noted above (§ 10.2.1), the status of such items, even when they are real words rather than synthesized ones, is questionable and they may be interpreted by bilinguals as borrowings (therefore easily assimilated to the base language) rather than switches. Second, switching studies are predicated on the idea that the language you start in affects the language you switch to. Yet if we admit that bilinguals can activate both languages simultaneously, a state surely to be achieved during CS or when accessing interlingual homophones, then we would expect that cross-linguistic interaction may operate bi-directionally (from base to guest or vice-versa). Third, switching studies, by their current design, cannot be informative regarding how long before or after a CS any cross-linguistic effect can be detected. In theory, it is possible that bilinguals adopt a bilingual production (or perceptual) mode, in which they may behave quite differently from when they expect to produce (or hear) in only one language. Given that bilinguals should not be assumed to perform to the phonetic norms of monolinguals, it is crucial to investigate the effects of CS relative to their own non-switching norms.

Linguistic studies devoted to describing the phonetic effects of CS, rather than the cognitive mechanisms underlying language switching, are few (Toribio et al. Reference Toribio, Bullock, Botero, Davis, Gess and Rubin2005; Bullock et al. Reference Bullock, Toribio, González, Dalola, O’Brien, Shea and Archibald2006; Khattab Reference Khattab, Hua and Dodd2006, this volume). Like switching studies, these have often induced CS in bilinguals in order to insure that the specific phonetic features under examination appear in the appropriate contexts with the difference that the materials used are intra-sentential CS constructions with grammatically constrained junctures, occurring either at the Subject–Predicate or Verb–Object boundaries. In this respect, the stimuli resemble natural bilingual CS. In laboratory studies of linguistic CS, researchers have attempted to redress the limitations imposed by the switching paradigm with respect to bilingual language production by posing additional questions, such as those cited in (5).

These research questions are cited here because they pose fundamental issues that any inquiry into bilingual CS should take into consideration. Item (5a) considers the general effect of CS on bilingual production because it is possible, in theory, that bilinguals manifest no difference between modes, or that they adopt compromised or merged phonetic values across a CS utterance relative to a monolingual one. Notice that questions (5b–c) raise the possibility that the effects of CS on phonetic production may be asymmetrical. That is, perhaps due to inherent linguistic differences or to speaker proficiency, to mention but a few factors, only one language of the pair may be affected (5b). Additionally, given that L1 phonetic values are assumed to be set early, it is possible that the language first acquired may be more stable during CS than the L2 (5c). Item (5d) aims to test the directionality assumption implicit in language switching studies and (5e) is designed to tease apart the effects of CS from that of language mode by examining whether perturbations to the phonetic system in CS are temporary or global.

Interestingly, the results of phonetic CS studies to date do not converge with those of the production switching studies reviewed above in § 10.3.1. In particular, the study by Bullock et al. (Reference Bullock, Toribio, González, Dalola, O’Brien, Shea and Archibald2006) showed a robust effect of CS on phonetic production that would not be predicted by a switching study. They tested the production of Spanish–English bilinguals in both monolingual and bilingual modes, separating their participants into two groups who were mismatched in proficiency. The Spanish (L1) bilinguals were strongly Spanish dominant and most had detectable foreign accents in their English. The English (L1) speakers, however, were Spanish instructors and, thus, more balanced across their languages. Each group was tested on their productions of /p,t,k/ in separate Spanish and English monolingual sessions. They were then tested in a bilingual session where they read CS sentences in both directions, randomly ordered. Embedded in each sentence were counterbalanced tokens of /p,t,k/ at strategic sites: pre-switch, at the switch juncture, and post-switch, as illustrated in (6).

The results showed that both groups, regardless of mode or switch site, maintain significantly distinct categories for Spanish versus English, a result that confirms the findings from the production switching studies. However, despite the participant group differences in L1 and in L2 proficiency, both groups showed an identical asymmetric pattern of phonetic shift in CS; that is, the effect of CS on production was manifested only in their English language productions. Specifically, their English language VOT values merged toward (but did not converge with) Spanish language values only when CS, but their Spanish language VOT productions remained constant across modes. The influence of Spanish on English occurred regardless of the direction of the switch. Intriguingly, the phonetic merger was most pronounced before switching from English to Spanish, rather than in the reverse direction. That is, bilinguals showed the highest degree of phonetic merger in anticipation of CS. When switching from Spanish to English, their English VOT productions at the switch site also merged significantly toward the Spanish language values while, at the post-switch position, they recovered their own monolingual values.

These findings suggest that there is a cross-linguistic effect in CS but one that is more complex than anticipated by switching studies. This effect appears to be local, rather than global, as it is concentrated before and directly after the switch. It is also independent of the base language (i.e. the language that you start in) because it occurred regardless of the direction of the switch. In fact, the English language productions of both groups were most Spanish-like when speakers began an utterance in English. Finally, the effect can be asymmetric, affecting only one language of the pair whether it is the base or the guest language. Importantly, the convergence between languages is not complete; these bilinguals, regardless of proficiency level, maintained separate voicing categories across their two languages, although not necessarily in the identical range to those of monolinguals of the respective languages.

The authors of the study speculate that the observed asymmetry may be due to inherent linguistic differences. That is, the VOT range for voiceless stops in English is expansive compared to the relatively compressed range of the short lag stops of Spanish. This could potentially allow more flexibility in the production of voiceless stops in English, permitting convergence toward (but not confusion with) Spanish during CS.¹ By contrast, expanding the VOT continuum of voiceless stops for Spanish past a certain interval (>30ms) may push them noticeably out of the Spanish range. This would suggest that inherent phonetic differences may condition CS behavior and, as within the morpho-syntactic domain, the output of CS must respect the phonological constraints of both languages, albeit allowing for phonetic variability in their expression.

Only one study to date examines directly whether CS can confound phonological distribution. Bullock et al. (Reference Bullock, Toribio, Davis, Botero, Chand, Kelleher, Rodríguez and Schmeiser2005) investigated whether CS could impact the production of syllable final lateral allophones among Puerto Rican Spanish (PRS)–American English (AE) bilinguals. Both languages possess phonological processes that impact syllable final liquids. In AE, a final lateral is produced with a retracted tongue dorsum and realized as a velarized, or dark l: [ɬ]. A salient (and sociolinguistically stigmatized) property of PRS is the variable application of lamdacization where an underlying rhotic surfaces as a lateral (e.g. vivir “to live” → [biβíl]). In PRS, syllable final laterals are apico-alveolar but they may surface as the reflex of either an underlying /l/ or of an underlying /r/. This means that the distributional as well as the phonetic properties of laterals differ between these two languages. The study was designed to test whether bilinguals could be observed to confuse the phonologies of their two languages by producing the alternate language allophone while engaged in reading CS sentences such as as in (7):

Extracting each lateral produced in both monolingual and CS contexts, the researchers measured the degree of velarization of all lateral productions by reference to the position of the second formant (F₂) – a velarized lateral will show a significantly lower F₂ than an apico-alveolar lateral (i.e. “clear l”). There was a small effect of CS within the Spanish language productions among individual speakers in that one speaker only produced lambdacization of underlying /r/ and another produced significantly velarized variants for underlying /l/ only while CS. The researchers suggest that, “it may be more difficult to . . . self-monitor pronunciation” while CS (Bullock et al. Reference Bullock, Toribio, Davis, Botero, Chand, Kelleher, Rodríguez and Schmeiser2005:110). However, overall, the results showed that these bilinguals, even in CS, maintain separate, correctly distributed allophones across their two languages. That is, they did not confuse their phonologies while engaged in CS.

In sum, laboratory studies investigating the effect of CS on production demonstrate that cross-linguistic influence is present at the phonetic level even though bilinguals are successful in maintaining separate phonological categories across languages. However, it also shows that the interplay between CS and phonetics is complex and may, in part, be determined by the specific phonetic properties under investigation.

10.5 Can phonology constrain CS?

Up to this point we have considered only whether CS affects phonological/phonetic structure. The issue can be viewed the other way around: can phonological/phonetic structure affect CS? This question is the natural corollary to the syntactic theoretic literature devoted to CS, yet only rarely has a role for phonology been acknowledged in the search for linguistic constraints on CS. The few proposals that exist view the role of phonology as facilitating, not constraining, CS and at a lexical rather than a phrasal level (Clyne Reference Clyne2003). The idea behind facilitation, as envisioned by Clyne (Reference Clyne2003), is that certain lexical items can act as triggers for CS in bilingual speech. Because there generally needs to be some similarity in the surface form of a trigger word across the component languages, facilitation is more likely to arise in closely related languages, but it is not unattested in typologically distinct languages.

According to Clyne, certain types of words – bilingual homophones, unassimilated lexical transfers (i.e. nonce borrowings), and proper names – may facilitate a shift in language, as illustrated in (8).

The bilingual homophone smal (Dutch “narrow”) has converged phonetically for the speaker cited in (8) and he pronounces it identically across Dutch and English: [smɑl]. The coincidence of the phonetic surface form across languages triggers a CS in an unlikely syntactic context (between a modifier and adjective). This implies that facilitation (triggering) can contravene syntactic constraints.

Facilitation has also been reported at the prosodic level in Vietnamese–English CS (Tuc Reference Tuc2003). Standard Vietnamese has a repertoire of six distinctive tones, each designated by a name and represented orthographically by a diacritic (or by the absence of a diacritic for the “neutral” tone ngang), as given in (9).

Tuc (Reference Tuc2003) shows that of these six tones, the last three, characterized by contours, by glottalization, or a combination of both, are virtually excluded from occurring immediately before CS into English.² The remaining tones have a relatively high or mid pitch, which Tuc argues facilitates switching into English because Vietnamese speakers establish a perceptual equivalence between the high and mid Vietnamese tones with the stressed and unstressed syllables of English, respectively. Thus, CS into English overwhelmingly occurs at the tonal range that is most appropriate for both languages. Zheng (Reference Zheng1997, cited in Clyne Reference Clyne2003) finds that switching between Mandarin and English is similarly restricted to a particular tonal range that is perceived to be compatible to both languages.

On another interpretation of these data, one could argue that the tonal properties of Vietnamese (and perhaps Mandarin) do more than facilitate a CS; they appear to constrain it. It is not simply the case that lexemes bearing particular tones trigger CS but CS is virtually blocked unless certain tones appear at a switch juncture. This can be seen when a particle with no syntactic function in an utterance is inserted before a CS, as in the example in (10), where the determiner đó “that” has been inserted.

As Tuc shows, the CS sentence would be fully grammatical without đó. In fact, the corresponding monolingual phrase would be ungrammatical if the determiner were to precede the equivalent Vietnamese verb for “recall.” But the presence of the dummy determiner, which Tuc endows with the pragmatic function of signaling CS, can be understood to be prosodically motivated. Without it, the sentence may be grammatical but the CS would likely be ill-formed as it would be directly preceded by a contour tone, rather than a mid or high tone. This implies that the particle is inserted not simply to facilitate CS but, instead, to allow it.

In sum, the data summarized in this section provide empirical support to the notion that CS can be conditioned by phonological structure. The Vietnamese data, in particular, strongly suggest that CS may be subject to prosodic constraints. This implies that although the search for structural constraints on CS has largely been confined to the morpho-syntactic component of grammar, it may be time to expand the quest to consider the role of prosody in CS. This is but one of a number of topics that awaits future study.

10.6 Conclusion: challenges for future research

This chapter began by laying out three general questions concerning the role of phonology and phonetics in CS. Here, we consider them in turn in an attempt to advance some conclusions.

1. (1.) Does CS have an effect on phonological/phonetic production and perception?

   This question probes whether two languages may overlap or influence one another in CS. As we have seen, there is clear evidence of crossover between languages in the production domain but only at the phonetic level; phonological categories do not appear to overlap in CS. In the perceptual domain, there is also reported evidence of cross-linguistic influence on the processing of acoustic stimuli. In particular, phonological dissimilarity between languages has been shown to have a facilitative effect on both perception and word recognition. The answer to question (1.), then, depends upon the degree and type of overlap concerned but, by and large, we do find effects of CS on both production and perception. Indeed, this is the expected result under models that assume that bilinguals maintain both languages simultaneously activated.

2. (2.) Can phonological/phonetic properties be observed to constrain CS production?

   Although constraints on CS have been the main preoccupation of syntacticians interested in bilingualism, this issue has been only cursorily addressed in the phonological literature. Results from studies of CS between languages with typologically distinct prosodic systems suggest that the answer to this question is affirmative. But, clearly, this is an area that merits much more consideration.

3. (3.) Is there a phonetic base or matrix language in CS?

   If the base language is construed to be the language that initiates a CS utterance then the answer to this question is negative. Phonetic overlap can occur irrespective of the direction of CS. However, there remains the possibility that some bilinguals may show a greater influence of one language over the other only in their CS pronunciations.

The answers to the questions posed in this chapter are only tentative and an exploration of the phonetic reflexes of CS remains very much an open field of inquiry. Clearly, the factors underlying phonetic production and perception in CS are complicated and present significant challenges to future researchers. Data collection, alone, is an obstacle because future work must undertake rigorous acoustic phonetic analyses of CS speech as impressionistic transcriptions are not detailed enough to detect the myriad cues that may be present. The addition of data from spontaneous CS corpora is an absolute necessity but, here, researchers will be hampered by the difficulty of collecting a sufficient amount of target tokens in the appropriate contexts.

A significant challenge to understanding the phonetics of CS arises from the fact that bilingual phonology, in general, is much understudied. Not all bilinguals are “accent free” in both languages and it is quite likely that even those who pass as monolinguals differ substantially from true monolinguals at the phonetic level. Thus, it is imperative that researchers examine bilinguals’ CS behavior in relation to the participants’ own monolingual performance. In addition, researchers repeatedly underscore the highly variable nature of bilingual phonetic performance; even bilinguals of virtually identical sociolinguistic profiles can behave quite differently at the phonetic level. Thus, group results must be treated with caution as they tend to efface the often dramatic differences that individuals may manifest in their speech production.

Finally, although there are many avenues of bilingual phonology to investigate in relation to CS, this chapter will conclude with one in particular. Neglected in much of the discussion regarding CS is the fact that phonology and syntax interface in bilingual performance. In this respect, a fruitful area for research in CS will likely be found at the prosodic level where pitch contours range across an utterance and, in many languages, are used for various discourse-pragmatic purposes. Prosodic or accentual boundaries are not necessarily isomorphic to syntactic or lexical ones. For instance, pitch peaks in many languages, like Spanish, are not bound to the stressed syllable of a lexical item but may, at times, be aligned to a syllable in the following word. Does this affect CS? How do bilinguals jointly reconcile the syntactic constraints of their component languages with the prosodic ones? These are unanswered questions but it seems quite likely that the interface between prosody and syntax, and not merely syntax alone, may play a role in circumscribing the domain of CS.

11.1 The search for universal constraints on code-switching

In the last few decades there has been a burgeoning growth in the literature of code-switching (hereafter CS), the use of two (or more) languages between sentences (i.e. inter-sentential) or within a sentence (i.e. intra-sentential). Whereas the pioneering works focused on bilingual communities in the United States (e.g. Poplack Reference Poplack1980), by now there are also studies on bilingual CS in different parts of the world, including Europe (e.g. Backus Reference Backus1996), Asia (e.g. Chan Reference Chan1998a, Reference Chan and Pennington1998b), Africa (e.g. Myers-Scotton Reference Myers-Scotton1993a, Reference Myers-Scotton1993b) and the Middle East (e.g. Berk-Seligson Reference Berk-Seligson1986). In terms of language typology, while Indo-European language pairs tend to attract attention, there have been additional studies on languages that come from families other than Indo-European. These efforts have been spurred by a growing interest in CS around the world and an increasing recognition of bilingualism as a proper sub-discipline of linguistics – the broader field to which CS belongs (Li Wei Reference Li Wei2000, Myers-Scotton Reference Myers-Scotton2006a, Romaine Reference Romaine1995). In addition, researchers have recognized that CS is not an indicator of deficiency in either or both language(s). Instead, it is most often viewed as a resource that bilinguals tactfully utilize to achieve various communicative effects (see Gardner-Chloros, this volume), to index social roles and identities (Myers-Scotton Reference Myers-Scotton1993b), and/or to manage ongoing talk (see Gafaranga, this volume). Furthermore, researchers have come to realize that (intra-sentential) CS – far from being random – is patterned and structurally governed, although there is still debate about the nature of grammatical constraints on CS and whether these constraints are universal (but see Myers-Scotton and Jake, MacSwan, this volume).

The search for universal constraints on CS has been instrumental in spawning much research on new data involving various language pairs. Based on Spanish–English data collected from Puerto Ricans in New York City, the classic paper by Poplack (Reference Poplack1980) proposed The Free Morpheme Constraint and The Equivalence Constraint, the two constraints that have probably been most frequently discussed in the literature. The Free Morpheme Constraint holds that CS does not take place within a word between a free morpheme and a bound morpheme (e.g. CS between “eat,” an English verb stem, and “-iendo,” the Spanish present progressive, is impossible (Poplack Reference Poplack1980:586) unless the former is phonologically integrated into Spanish (Sankoff and Poplack Reference Sankoff and Shana1981). The Equivalence Constraint stipulates that CS only takes place where the surface order of constituents surrounding the switch point is the same in the participating languages (e.g. CS is possible between an adjective and a noun if the participating languages both have the same pre-nominal (or post-nominal) positioning of adjectives (Pfaff Reference Pfaff1979)).

Poplack (Reference Poplack1980) concluded that the two constraints apply to her Spanish–English data irrespective of L2 proficiency and sociolinguistic variation; Sankoff and Poplack (Reference Sankoff and Shana1981:7) further suggested that the constraints might well be universal. This claim was the impetus for testing the validity of these constraints on other language pairs, which often resulted in revised or new constraints. Nartey (Reference Nartey1982) was quick to point out that data of Adaŋme–English (spoken by educated Ghanaians in Ghana) present counter-examples to both constraints, and she was probably the first one to suggest that different linguistic constraints may apply in different socio-cultural environments – a precursor of more recent works such as Bhatt (Reference Bhatt1997) and Muysken (Reference Muysken2000). Much as data from Spanish–Hebrew collected in Jerusalem (Berk-Seligson Reference Berk-Seligson1986) and Arabic–French collected in Morocco (Bentahila and Davies Reference Bentahila and Davies1983) supported The Free Morpheme Constraint, counter-examples were found against The Equivalence Constraint. Bentahila and Davies (Reference Bentahila and Davies1983) suggested that subcategorization restrictions (or selection) – rather than word order equivalence – are always respected in CS. Di Sciullo et al. (Reference Di Sciullo, Muysken and Singh1986) refuted The Equivalence Constraint for reasons which are empirical (i.e. many possible switching sites allowed by The Equivalence Constraint show little CS in Poplack’s dataset) and theoretical (i.e. The Equivalence Constraint does not refer to deeper structural relations). They examined data from Italian–French–English CS and Hindi–English CS, and put forth The Government Constraint, which bars CS between a lexical head (e.g. a verb) and the “highest” element in the constituent this head governs (e.g. the determiner in the verb’s object noun phrase). The Government Constraint was found to be empirically inadequate by Belazi et al. (Reference Belazi, Rubin and Toribio1994), among others. These authors propose The Functional Head Constraint, which precludes CS between a functional head (i.e. Determiner, Inflection, Complementizer, Quantifier and Negation) and its complement (e.g. Noun Phrase, Verb Phrase, Inflection Phrase), which the functional head f-selects (see Abney Reference Abney1987). Again, counter-examples were soon brought to light by Bhatt (Reference Bhatt, Fuller, Han and Parkinson1995), Halmari (Reference Halmari1997) and Mahootian and Santorini (Reference Mahootian and Santorini1996) from various language pairs documented in previous literature.

Poplack and her associates also examined CS in more exotic language pairs, including Finnish–English (Poplack et al. Reference Poplack, Wheeler and Westwood1989), Tamil–English (Sankoff et al. Reference Sankoff, Poplack and Vanniarajan1990), Wolof–French and Fongbe–French (Poplack and Meechan Reference Poplack, Marjory, Milroy and Muysken1995). Facing apparent counter-examples to The Free Morpheme Constraint and The Equivalence Constraint, they concluded that these instances are in fact “nonce borrowings” rather than CS. The suggestion generated a series of debates as to the definition and delineation between CS with respect to (nonce-)borrowing (Myers-Scotton Reference Myers-Scotton1993b, Reference Myers-Scotton2002a; Muysken Reference Muysken2000). In any case, Poplack’s original constraints have not faded away in current literature. The empirical predictions of The Free Morpheme Constraint have been preserved in the PF (Phonetic Form) Adjunction Theorem of MacSwan (Reference MacSwan1999a, Reference MacSwan1999b, Reference MacSwan2000), based on Nahuatl–Spanish data, whereas The Equivalence Constraint is still seen as facilitating CS (Muysken Reference Muysken2000) if not strictly constraining it (Deuchar Reference Deuchar2005, based on Welsh–English data).

In many datasets there is an obvious asymmetry between the participating languages in terms of the morpho-syntax of CS sentences. Most of these data involve an Indo-European language and an Asian or African language (e.g. Kamwangamalu Reference Kamwangamalu, Music, Graczyk and Wiltshire1989 on Bantu–English/French; Sridhar and Sridhar Reference Sridhar and Sridhar1980 on Kannada–English; Nishimura Reference Nishimura1985a, Reference Nishimura and Vaid1985b on Japanese–English; Park Reference Park1990 on Korean–English; Joshi Reference Joshi, Dowty, Karttunen and Zwicky1985a, Reference Joshi, Dowty, Karttunen and Zwicky1985b on Marathi–English in which both languages are Indo-European). All these works eventually paved the way to the Matrix Language Frame Model (henceforth the MLF Model) of Myers-Scotton (Reference Myers-Scotton1993a, Reference Myers-Scotton1997, Reference Myers-Scotton2002a), which drew support primarily from Swahili–English CS data. This remains a dominant paradigm in the grammatical approach to CS. The basic premise of the MLF Model is that in a code-switched “sentence,” defined as a Complementizer Phrase (CP) by Myers-Scotton, the Matrix Language (ML) generates the sentence structure. This implies two things. First, the ML determines word order, a constraint formalized as The Morpheme Order Principle, and, second, the ML supplies system morphemes (largely bound morphemes and function words), a constraint formalized as The System Morpheme Principle (see Myers-Scotton and Jake, this volume). The Embedded Language (EL), on the other hand, can only contribute content morphemes, i.e. content words that take part in theta-marking such as nouns, verbs, adjectives, and most prepositions. In addition to these two core principles, there are a number of subsidiary principles that deal with alleged counter-examples, mostly cases involving system morphemes from the EL. For instance, the Embedded Language Island Principle allows EL system morphemes to appear in EL phrases consisting of all words from the EL. The Double Morphology Principle licenses an EL system morpheme (e.g. a plural morpheme) if it is “doubled” with its counter-part from the ML. Apart from positing subsidiary principles, the content/system morpheme distinction has also been fine-tuned to account for apparent counter-examples to the two overarching principles and now there are four types of morphemes rather than the original two. In any case, skeptics may still be dissatisfied with the subsidiary principles and the new morpheme models that water down the original force of The System Morpheme Principle. Also, the additions have made the MLF model perhaps too sophisticated and uneconomical to be desirable as a model of bilingual competence (Chomsky Reference Chomsky1965), in marked contrast with a more recent view that there are actually no constraints or principles specific to CS in the language faculty of bilinguals. This alternative is known as The Null Theory (see MacSwan Reference MacSwan1999a, Reference MacSwan1999b, this volume; Mahootian Reference Mahootian1993; Chan Reference Chan2003).

Indeed, many constraints and models in the CS literature have been challenged and modified by the discovery of new empirical data. Unlike linguistic research with monolingual participants, it is unclear whether a bilingual’s intuitions of CS sentences are consistent and trustworthy. For one thing, grammaticality judgments on hypothetical CS sentences may be affected by the social stigma that has always been attached to CS (Pfaff Reference Pfaff1979). A more deep-rooted problem lies in the fact that many bilinguals are non-balanced, i.e. they have not attained a proficiency level in their weaker language that is akin to that of a native speaker of that language. In this respect, some researchers have been able to solicit consistent grammatical judgments (Bentahila and Davies Reference Bentahila and Davies1983), while others have found varied intuitions among different proficiency groups (Toribio Reference Toribio2001b). Still others have been most careful to elicit judgments only from balanced bilinguals (MacSwan Reference MacSwan1999a, Reference MacSwan1999b), only to leave aside the copious data produced by non-balanced bilinguals, whereas others avoided grammaticality judgments altogether (Mahootian Reference Mahootian1993).

Given that consistent and trustworthy judgments are not always obtainable and that CS data display considerable variation, it is understandable that some researchers have become disgruntled with the constraint approach to CS. Nonetheless, one must be careful in differentiating three different responses:

1. (1.) The constraint approach to CS is misguided and futile; it does not provide any insights into CS (see Bokamba Reference Bokamba1989; Gardner-Chloros and Edwards Reference Gardner-Chloros and Edwards2004).

2. (2.) There are no universal constraints on CS, but there are specific constraints which are followed in different bilingual communities (see Bhatt Reference Bhatt1997; Muysken Reference Muysken2000).

3. (3.) There are no constraints that operate specifically on CS. It is governed by abstract constraints or principles underlying universal grammar on a par with “pure” languages (see Mahootian Reference Mahootian1993; MacSwan Reference MacSwan1999a, Reference MacSwan1999b, Reference MacSwan2000; Chan Reference Chan2003).

Position (1.) downplays the fact that, although universal constraints are not yet in sight, CS does exhibit certain structural regularities, patterns that led researchers to arrive at various constraints or models. CS is often produced fluently and understood instantly by bilinguals in the manner of spontaneous monolingual conversations, and the CS sentences do show structure; they are not a loose array of words glued together randomly, nor do they look similar to those sentences produced by severe agrammatic aphasics. Even linguists who have dismissed the constraint approach have never denied that there is syntactic structure underlying CS; rather, they prefer accounts of CS patterns based on social–pragmatic factors instead of syntactic theory.

Position (2.) is exemplified by Bhatt (Reference Bhatt1997), who adheres to grammatical constraints, but not those that are absolutely inviolable. Rather, he explains variation in CS patterns by different rankings of constraints, as envisioned in Optimality Theory. As a particular language pair is assigned a certain constraint ranking, there remains a problem of how to capture variation within a language pair. Muysken (Reference Muysken2000) does not employ Optimality Theory, but he attempts to accommodate various CS patterns by devising a typology of language mixing. In particular, a bilingual may exploit one of these three strategies, namely, alternation, insertion or congruent lexicalization when engaging in CS. Alternation refers to a “total” switch to another language, including lexicon and grammar. In insertion, bilinguals do not switch totally; they insert words and phrases from one language (i.e. the EL) into a sentence frame generated by the grammar of another language (i.e. the Matrix Language). Congruent lexicalization applies to a language pair with languages that are typologically related (e.g. Dutch and German); the structure of a CS sentence is very similar to that found in both participating languages, to the extent that CS apparently can take place at any point of the sentence with words drawn from either lexicon. The ingenuity of Muysken’s (Reference Muysken2000) model lies in its attempt to connect the syntax of CS with sociolinguistics and psycholinguistics. He suggests that alternation is typical in stable bilingual communities with balanced bilinguals, whereas insertion is commonly found in former colonial settings where bilinguals are more fluent and at home with their first language. Congruent lexicalization is found where bilinguals, often second-generation immigrants, are fluent in two typologically similar languages, and these languages have equal prestige in society. Yet, syntax still plays a crucial role in this model. The three strategies represent the limits within which CS may vary on an individual or community level, although the three strategies are themselves defined by diagnostic criteria which are syntactic.

Position (3.) is associated with The Null Theory (Mahootian Reference Mahootian1993; MacSwan Reference MacSwan1999a, Reference MacSwan1999b; Chan Reference Chan2003). This theory considers that CS is possible as long as no principle or constraint in universal grammar is violated, and therefore it departs from most previous models or constraints that are supposed to operate on CS specifically. The idea is motivated by Occam’s Razor – in an attempt to devise the most economical model of CS (Mahootian Reference Mahootian1993; MacSwan Reference MacSwan1999a, Reference MacSwan1999b, Reference MacSwan2000), as well as by a consideration of cognitive economy (Chan Reference Chan2003). It would be more parsimonious for a bilingual mind not to have a separate grammar for CS. That is, considering that only some speakers in the world are bilinguals who may engage in CS, it is uneconomical to propose a code-switching grammar within the language faculty or universal grammar, which is supposed to be innate, inborn, and biologically endowed in all human beings, bilingual or monolingual. And even though there may be a CS grammar, it still begs an explanation as to why the putative constraints (e.g. The Free Morpheme Constraint, The Equivalence Constraint, the MLF Model, etc.) look so radically different from principles or constraints in other languages (e.g. structural dependence, locality, binding principles).

The major drawback of The Null Theory is that constraints are necessarily expressed in theory-specific terms, and therefore the empirical predictions may vary when researchers are committed to different linguistic theories or assumptions.¹ One may say that The Null Theory is more of a spirit rather than a coherent framework, treating CS on a par with other “pure” languages rather than as a “peripheral” phenomenon, and applying to CS various syntactic theories which are independently grounded in monolingual grammatical phenomena. Despite the thrust of The Null Theory, it is in this connection that it meets with another problem – there are a number of CS constructions which do seem to arise specifically through language contact and which are absent in either participating language, including the mixed compound verbs (i.e. a code-switched verb and a helping verb from the host or matrix language – see below), the portmanteau construction (e.g. a “mirror” sentence in which a verb from a VO language co-occurs with another one from an OV language – see below), and the omission or double marking of function words or bound morphemes (e.g. the omission of determiner(s) – see below). To explain these phenomena, general syntactic rules or principles (e.g. word order parameters, agreement, etc.), which The Null Theory appeals to, seem insufficient, but additional, construction-specific rules or principles, apart from those of the two participating languages, appear necessary.

The above survey hardly solves the intricate empirical and theoretical problems involved in the grammatical study of CS, but merely highlights the main dilemmas that researchers have been facing. On the one hand, CS data often turn out to be more diverse and varied than the proposed models or constraints would have predicted. On the other hand, there is structural regularity underlying most CS sentences, which requires an explanation. The tension between constraining CS by grammatical theories and describing the diversity of real data is indeed daunting. A related issue is whether to account for the various CS patterns in sociolinguistic or pragmatic terms (which are more concerned with variation) or by syntactic theories (which are more concerned with underlying regularity). A third dilemma concerns adopting a uniform, universalist theory (e.g. the MLF Model), which is supposed to apply in all language pairs, or a typological approach (e.g. the Bilingual Speech Model of Muysken Reference Muysken2000), which posits different constraints in different bilingual communities. Yet a fourth one is whether to devise constraints that operate on CS specifically, or to resort to syntactic theories independently grounded in monolingual grammatical phenomena.

One way of resolving the above dilemmas and synthesizing the insights of various approaches is to envision a greater role for processing strategies, conceived as part of linguistic performance in the generative literature (Chomsky Reference Chomsky1965). Performance is also the site where pragmatic, discourse-functional and other sociolinguistic factors may come into play and interact with grammar. This is consistent with the typologists’ view that processing strategies and other language-external factors may shape grammars and grammatical constructions (Comrie Reference Comrie1989; Croft Reference Croft2003). In the context of CS, a bilingual gets to know not only words but also syntactic rules from both languages. Should the two languages have different rules for a construction (e.g. a noun phrase or a verb phrase), the bilingual has access to more than one syntactically viable option in forming a CS construction. Various options are taken in different bilingual communities, hence the diversity of patterns. The following sections illustrate how such an approach can be applied to two cases where CS involves typologically different languages.

11.2 Code-switching between VO and OV languages

One area where languages vary from each other syntactically is word order. Typologists have long treated subject–verb–object order as a major parameter by which to classify languages. In those cases where a bilingual code-switches between a VO language and an OV language, what patterns are produced? There are four logical possibilities:

Most researchers have believed that word order follows the language of the verb; in other words, VO order is attested if the verb comes from the VO language (i.e. pattern (1.)), whereas OV order is realized if the verb comes from the OV language (i.e. pattern (2.)). The following are some examples found in the literature.

Patterns (3.) and (4.) appear to be ruled out. Various grammatical theories have been invoked to explain this apparent restriction. Mahootian (Reference Mahootian1993) appeals to Tree Adjoining Grammar in which the lexical content of a verb is specified for its arguments (i.e. subject and object of the verb) as well as the position of these arguments. MacSwan (Reference MacSwan1999a, Reference MacSwan1999b) proposes that a verb carries a case feature that derives VO or OV order in a Minimalist fashion. Nishimura and Yoon (Reference Nishimura, Keumsil Kim and Silva1998) suggest that the directionality of the head (verb in this case) has to be followed. In spite of different explanatory tools, all of these approaches assume a lexicalist account of VO/OV order – that the latter is in some way specified by the head verb.

One may wonder whether this restriction is universal. For one thing, The MLF Model would have allowed that a bilingual inserts a verb, a canonical content morpheme, from a VO language into an OV order, or insert a verb from an OV language into a VO order, provided that the Matrix Language is OV or VO respectively. A deeper probe into the literature indeed yields data where verb–object order does not follow the language of the verb (Chan Reference Chan2003, Reference Chan2008), as shown here:

The mixed compound verbs further attest pattern (4.) in which apparently a verb from a VO language assumes OV order, for instance, in the Matathi–English example in (5).

There has been quite some dispute as to the proper status of the code-switched verb, though. As the name “mixed compound verb” suggests, some researchers consider that the code-switched verb (from the VO language) is actually adjoined to the helping verb from the OV language (e.g. kartāt in (5)); in other words, the code-switched element and the helping verb together form a V node. The helping verb somehow “nativizes” the code-switched element (Kachru Reference Kachru1978), and so the “mixed compound verb” functions as any other verb in the OV language, hence the OV order. The code-switched element may be a verb or a nominalization. There is, however, evidence that, at least in some cases, the code-switched element is a verb on its own. For instance, in (6) below, the English verb (“force”) takes a Panjabi object (“baceã”) inflected by an accusative marker (“nũ”). The helping verb in Hindi (“kər”) most probably rests in a higher syntactic position, since it is separated from the English verb (“force”) by a negation marker (“nəi”) (see Muysken Reference Muysken2000 for more discussion on these “bilingual verbs”).

The existence of examples in (3), (4), and (6) (i.e. patterns (3.) and (4.) above) resist a lexicalist account of verb–object order in CS, where verb–object order is specified in the head verb. Alternatively, VO or OV order could be stipulated by a syntactic rule as more traditionally envisioned, for instance, a verb-initial and a verb-final parameter (Neeleman and Weerman Reference Neeleman and Fred1999, Saito and Fukui Reference Saito and Naoki1998). Now, suppose the bilingual has access to both rules. Either VO or OV would comply with the input and thus both orders can be found in the production data. On the other hand, if the bilingual’s languages both have VO order (e.g. Spanish–English, Cantonese–English), the bilingual does not have the OV option and OV order would never be yielded, unless the objects are moved or pre-posed for pragmatic effects. By the same token, it is hypothesized that CS between two OV languages does not yield VO order unless it is for “pragmatic” effects.

Consider again the bilingual who code-switches between a VO and an OV language. The idea that either VO or OV is an “optimal” strategy may sound like a mere description of the above data, but it is not: Both VO and OV rules may be co-activated, leading to “portmanteau” constructions such as those in (7).

The markedness of “portmanteau” constructions in comparison with VO or OV order (Backus Reference Backus1996; Nishimura Reference Nishimura1985a; Sankoff et al. Reference Sankoff, Poplack and Vanniarajan1990) may well be due to the relative economy of the latter option. In portmanteau constructions semantic information is duplicated in two verbs: one from the VO language and another from the OV language, while the selection of VO or OV order is non-redundant.

There is also a third possible option: neither VO nor OV order. It is difficult to conceive how this option can be realized if both verb and object are coded, in consideration of the fact that all sentences need to be linearized, a requirement which is imposed by the PF (Phonetic Form) Interface in the Minimalist Program (Chomsky Reference Chomsky1995). Another possibility is that the whole VP is not realized at all (hence neither VO nor OV), thus best respecting economy. Nonetheless, this would violate syntagmatic isomorphism – the principle that a meaning is expressed by a linguistic form (Croft Reference Croft2003; Haiman Reference Haiman1980). Since the verb also licenses the subject argument, the non-coding of VP implies the non-coding of the whole proposition. Yet another possibility is that the object is not expressed, which again respects economy but violates isomorphism. Empirically, there are indeed some instances where the object of a CS verb is dropped, such as the English–Japanese example in (8) below. However, Nishimura (Reference Nishimura1985a:138) attributed the object drop in (8) to the grammar of Japanese in which an object may be null where it is salient in the context. In this light, it is unlikely that object drop is induced by CS between a VO and an OV language, which is not attested elsewhere in the literature (see examples in (1), (2), (3) and (4) above).

Table 11.1 below summarizes the options of the processor for a bilingual who code-switches between a VO language and an OV language.

Table 11.1 Code-switching between a VO and an OV language: options of the processor

Portmanteau constructions are not ruled out strictly as ungrammatical. The sociolinguistic or pragmatic contexts where portmanteau constructions appear have not been adequately studied. However, the available data are often documented from immigrants, mostly of the second or intermediate generation (Nishimura Reference Nishimura1985a; Backus Reference Backus1996). They tend to engage in extensive conversational CS within their peer groups, and they stand midway between two different identities (i.e. those of their origin and of the new settlement). It may be the case that the portmanteau constructions, in juxtaposing the lexicons and observing the grammars of both languages, are a convenient strategy to signal these bilinguals’ dual identity or to involve mixed audiences of different groups (Nishimura Reference Nishimura1995). If this is on the right track, the portmanteau constructions would have been justified by sociolinguistic reasons even though they are not optimal. The least optimal strategy – i.e. neither VO nor OV – is theoretically possible but it may be syntactically impossible (where both verb and object are coded without a recognizable order, hence violating linearization) or impossible to prove (where both verb and object are not coded, hence violating isomorphism). Object drop appears to be permissible only where one of the participating languages (e.g. Japanese in (8)) licenses null objects in discourse contexts where the objects are readily recoverable, hence not clearly induced by CS. The apparent impossibility of CS inducing object drop may be due to other grammatical principles.

How does the bilingual select VO or OV order out of the two alternatives? There are two main approaches: the syntactic approach and the processing approach. In the former, some built-in syntactic mechanism derives the preferred OV/VO order. Under one view, consistent with Kayne’s (Reference Kayne1994) Linear Correspondence Axiom, VO is the unmarked, base-generated order, whereas OV is derived through movement of the object to a higher position. As for the processing approach (Chan Reference Chan2003, Reference Chan2008), both VO and OV can be considered to be base-generated word orders, selected by the processor under the influence of various “usage” factors. For instance, a particular order (OV or VO) has been in use and much more accessible, possibly because the speakers have been primed by using that language. The choice of one particular construction (e.g. VO or OV) is conventionally seen as governed by sociolinguistic norms (e.g. speakers, formality of occasion).

11.3 Code-switching between languages with different types of DP

Apart from word order, languages systematically vary from each other in terms of morpho-syntactic marking, functional words, and bound morphemes that encode grammatical information such as definiteness, tense, aspect, number, case, gender. In generative grammar, it has widely been assumed since Abney (Reference Abney1987) that noun phrases are actually determiner phrases (DP) headed by the determiner, although by “determiner” Abney had in mind various function words attached to nouns (e.g. articles, quantifiers, demonstratives). More recently, it has been recognized that the “functional domain” of nouns is more articulated than a Determiner (D) head (Cheng and Sybesma Reference Cheng and Rint1999; Giusti Reference Giusti and Haegeman1997; Ritter Reference Ritter1995). There are other functional heads hosting function words and even noun affixes, such as Quantifier (Q), Number (Num), Classifier (CL) and Case (K), and the functional domain may vary greatly from one language to another. For instance, whereas a referential argument (i.e. a DP that receives a theta-role in subject or object position) in English is mostly marked by an article, in Chinese languages, it is normally marked by a classifier which primarily expresses its attribute (e.g. shape), unit (e.g. similar to “cup” as in “a cup of tea”), or quantity (e.g. similar to “dozen” as in “a dozen cakes”). In other languages with morphological case (e.g. Hindi, Japanese, Korean, Turkish), an argument is always framed by a case marker that signals its role in an event (e.g. a nominative marked subject, an accusative marked object), although the case marker may be covert. For the sake of exposition, this difference can be captured by proposing that English DPs are Article Phrases (ArtP), whereas Chinese languages project Classifier Phrases (CLP) and languages with morphological case markers project Case Phrases (KP).

When there is CS between languages in which nouns are expanded into different functional projections, it seems that one functional template of the participating languages is chosen (i.e. ArtP, CLP or KP), as shown in (9) through (11).

The scenario is similar to that of the bilingual’s selecting either the VO or OV rule as elaborated in § 11.2 above. That is, CLP is projected instead of ArtP in (9), although the bilingual speaker is supposed to have access to the latter template as well. In a similar fashion, KP is projected instead of ArtP in (10), and ArtP is projected rather than KP in (11).

The plural morpheme in English (e.g. the English “-s”) is more likely to be expressed on a CS noun, when plurality is also expressed with determiners from another language (e.g. (12)) or is “double-marked” with another plural morpheme (e.g. ma-game-s (13) below). Myers-Scotton (Reference Myers-Scotton2002a, also Myers-Scotton and Jake, this volume) has explained these phenomena by stating that the plural morpheme is an “early system morpheme,” easily activated with the code-switched noun even though it may be from the EL.

Notice, however, that the DP in (12) is essentially a Classifier Phrase (CLP), the functional template from Cantonese. The plural morpheme arguably does not expand into a functional projection; that is, the plural noun “chapters” remains a noun (N) and hence it can be selected by the Cantonese classifier (Chan Reference Chan and Pennington1998b, Reference Chan2003). All in all, the Cantonese CLP template, as indicated by the numeral and classifier “gei2 go3/several,” is chosen even though English plural morphology is activated due to the context.

Sometimes, determiners from both languages are used to frame a noun, but the pattern appears to be specific to certain language pairs, for example, Moroccan-Arabic and French, as in (14) below.

Despite the co-occurrence of determiners in Arabic and French, apparently it is the Arabic functional domain rather than the French one that is selected in (14). Bentahila and Davies (Reference Bentahila and Davies1983) explained that in Arabic demonstratives and numerals select a DP – a noun with the definite article – and so in CS they also select a French DP with an article. A complementary point is that the French articles, unlike those in English and other languages, are always present with French nouns. The French article may easily be activated together with the French noun, which is similar to what the English plural morpheme is to an English noun in CS (see (12) and (13) above). Elsewhere, the use of determiners from both languages appears to be rare. The projection of functional templates from both languages is also rare and motivated by meaning. For instance, in (15) below the Korean case marker is required by Korean grammar, and the English definite article is motivated by the “superlative” meaning.

Sometimes, code-switched nouns may be bare, without determiners or affixes. Nonetheless, a number of these cases are explained by the grammar of the Matrix Language which licenses null determiners, especially in indefinite, generic, or predicative contexts, such as in (16), (17), and (18) below. In other words, the English nouns may well project to a CLP with a null classifier in (16) and (17), whereas in (18) the English noun projects to a KP with a null case particle.

It is hard, if not impossible, to find bare nouns in syntactic environments where the grammar of either language would have required determiners and noun affixes – Berk-Seligson (Reference Berk-Seligson1986) calls them code-switching “errors” – see (19). The functional template of neither of the participating languages is coded. This does not seem to be the normal case in and across language pairs.

Overall, the case of morpho-syntactic marking of nouns in CS described in this section is parallel to that of VO/OV order expounded in § 11.2 above, and hence the two may receive a unified account. When a bilingual code-switches between two languages with different types of DP, he or she has access to two morpho-syntactic rules with which to project the DP. The strategy is to activate one of the rules, probably as a “balance” of economy and isomorphism, but in some cases both may be activated, resulting in double marking, or neither may be activated, resulting in bare nouns without determiners. Double marking or bare nouns are more marked in terms of their occurrences in and across different language pairs.

Table 11.2 Code-switching between languages with different types of DP: options of the processor

11.4 Summary and areas for further research

This chapter has surveyed the major constraints and models that have been proposed to account for the structure of CS, highlighting the main controversies within the grammatical approach. The long-standing problem of these attempts has been the diversity and variation of CS patterns, which appear to defy any economical, unified, and universal syntactic account. It has been suggested that this diversity be construed as the interaction of the grammars of the two languages, which gives rise to more than one syntactically viable option. The selection of one of these options is the result of the influence of processing strategies and various sociolinguistic factors.

Despite the diversity of CS patterns, there seems a universal tendency to select one morpho-syntactic rule, probably a result of the balance between “economy” and “isomorphism” (Haiman Reference Haiman1980, Reference Haiman1983; Croft Reference Croft2003), which are functional principles rather than formal syntactic constraints. Less optimal patterns may appear when there is a functional motivation (sociolinguistic, pragmatic or processing), a hypothesis that opens avenues for further research. It might well be most efficient to select the morpho-syntactic rules consistently from one language, and that language would become what is conceived as the Matrix Language in the MLF Model. If this were correct, the Matrix Language would be an epiphenomenon, arising from language use or performance rather than competence. Then, theoretically, the series of principles that stipulate the role of the Matrix Language and the Embedded Language would be unnecessary. In other words, data that comply with the MLF Model represent “optimal” strategies under certain settings, but there may well be other strategies activating grammatical rules from both languages, when these choices are functionally motivated by sociolinguistic, pragmatic, or processing factors.

12.1 Introduction

Code-switching (hereafter CS) is a very common feature in the speech of bilinguals and has attracted the attention of sociolinguists and psycholinguists for many years. In the last forty years, syntacticians have become interested in the study of the structural aspects of CS. A wide range of (typologically different or similar) language pairs has been studied in order to find out whether constraints, both pragmatic and syntactic in nature, govern this speech style. A distinction has been made between inter-sentential and intra-sentential CS. The examples in (1) demonstrate inter-sentential CS in children, taken from Cantone and Müller (Reference Cantone and Natascha2005:210). In these examples, the bilingual German–Italian child Aurelio (age 2;5,21 in years;months,days) switches into Italian while speaking to a German interlocutor.

In the study of structural constraints in CS, researchers have mainly focused on intra-sentential mixes. Intra-sentential CS is the juxtaposition of elements from two (or more) languages within one sentence. In (2), each child switches into the respective other language within the same utterance, like Céline, a French–German bilingual child at age 2;11,15 in the French recording, and Leo, a Spanish–English bilingual child, at age 2;7 (see Liceras et al. Reference Liceras, Spradlin and Fernández Fuertes2005:239).

As for the alternation of both languages in child speech, researchers often use the term “code-mixing,” referring to the very early stages of bilingual first language development. The use of the term code-mixing reflects the idea that the alternation of the languages is not yet constrained (see Meisel Reference Meisel1994; Köppe and Meisel Reference Köppe, Meisel, Milroy and Muysken1995).

In the following sections, this chapter summarizes the most important constraints proposed for adult CS (§12.2) and the main studies on child language mixing (§12.3). §12.4 presents a radically monolingual view of adult CS, which implies that nothing, apart from the two grammatical systems involved, constrains adult CS (see MacSwan, this volume). This view applies to child code-mixing which, as a result, can be analyzed in the same way as adult CS.

12.2 Code-switching as a rule-governed behavior of bilingual speakers

12.3 Code-switching in bilingual children

12.3.3 The relation between language mixing and cross-linguistic influence

In 1978, Volterra and Taeschner published a longitudinal study on the simultaneous development of two languages (German and Italian) in two girls raised in Italy. Their view on simultaneous bilingualism is known as the three-stage-model. In this model, bilingual children necessarily pass through each of three stages until they reach the point of being truly bilingual, as illustrated in Figure 12.1.

Figure 12.1 The Three-Stage-Model

The first stage in development is characterized by the existence of one lexicon and one syntactic system (or grammar). This does not necessarily mean that the children use the grammar of one language in understanding and producing two languages; instead the syntactic system used by the children is best seen as a fused system, with properties from both languages. The view of a fused syntactic system will be illustrated further below. In general, the existence of one lexicon means that a bilingual child will not refer to a book as, for example, both buch (German) and libro (Italian) but will use only one of these words when producing in either language. From Volterra and Taeschner’s perspective, the results are mixed utterances of the type buch rotto “book_German broken_Italian” – buch voglio “book_German (I) want_Italian,” if buch is the word stored in the child’s lexicon.

The second stage is characterized by the development of two language-specific lexicons while there is still one syntactic system shared by both languages. As a result, the children use one grammar while drawing on two language-specific lexicons. In other words, the mixing between the two languages observed during the first stage (buch rotto and buch voglio) ceases to occur. Volterra and Taeschner present several examples of a fused grammatical system. For example, both children place attributive adjectives post-nominally, which gives target-like results for Italian, e.g. scarpe marrone scuro, “shoes brown dark” and il riso buono “the rice good,” but target-deviant placements in the German equivalents – schuhe dunkelbraun and reis gut.

During the third stage, the bilingual children, in addition to two separate lexicons, possess two separate syntactic systems. The view of Volterra and Taeschner that language mixing is due to a fused lexicon has been criticized, particularly since the seminal work of Genesee (Reference Genesee1989). Genesee demonstrates that there are other reasons for language mixing at the lexical level than the absence of an equivalent word in the lexicon of the respective other language. That his criticism is indeed plausible is indicated, among other things, by the observation that there are bilingual children who mix a lot in both languages, others only in one language, and still other children who do not mix at all or very rarely. One may thus conclude that code-mixing cannot be considered as evidence in favor of one, fused lexicon (see also Cantone and Müller Reference Cantone and Natascha2005).

Is language mixing in young bilingual children an instance of cross-linguistic influence? Cross-linguistic influence can be defined very generally by the influence of one language on the other (see Müller et al. Reference Müller, Kupisch, Schmitz and Cantone2006). Hulk and Müller (Reference Hulk and Müller2000) and Müller and Hulk (Reference Müller and Aafke2001) describe particular grammatical domains that are vulnerable for cross-linguistic influence in bilingual children. They propose conditions under which cross-linguistic influence affects bilingual first language acquisition. The interesting conclusion is that cross-linguistic influence seems to be determined only by linguistic properties of the grammatical phenomenon in question. The outcomes of cross-linguistic influence are acceleration (speed of acquisition is accelerated in language A due to the influence of language B), delay (speed of acquisition is reduced in language A due to the influence of language B), and transfer (the syntactic analysis of language A is used while speaking language B). With respect to language mixing in children, it seems that nothing, except the fact that lexical material from both languages has been used, will indicate that the two languages are influencing each other in the bilingual child. The mixed utterances produced by the children are well-formed from a syntactic point of view. Language mixing therefore has to be viewed and analyzed as a bilingual phenomenon that can occur independently of cross-linguistic influence, as evidenced in monolingual utterances.

In a study on gender marking in mixed DPs, i.e. in DPs that contain a determiner from language A and a noun from language B, Cantone and Müller (2008) find that German–Italian bilingual children respect the lexical gender of the noun when gender is marked on the determiner. This result is particularly revealing when the equivalents of these nouns have a different gender in the other language. For example, in (12a) Carlotta uses the Italian feminine noun pentola with a German indefinite article that carries the feminine suffix –e in the German context. The German equivalent of pentola is topf, which, however, is a masculine noun.

GC = German context, IC = Italian context.

(12)

(12)

1. (a.) eine_fem pentola_fem (Carlotta, GC, 2;9,25)

   a pot

   cf. German ein_mas topf_mas

2. (b.) il_mas schwanz_mas (Jan, IC, 3;7,1)

   the tail

   cf. Italian la_fem coda_fem

3. (c.) il_mas mio_mas stuhl_mas (Carlotta, IC, 2;7,13)

   the my chair

   cf. Italian la_fem mia_fem sedia_fem

4. (d.) una_fem blume_fem (Aurelio, IC, 3;8,13)

   a flower

   cf. Italian un_mas fiore_mas

5. (e.) für die_fem cintura_fem (Carlotta, GC, 3;6,3)

   for the belt

   cf. German für_mas den_mas gurt_mas

6. (f.) un_mas stern_mas (Lukas, IC, 3;5,8)

   a star

   cf. Italian la_fem stella_fem

If the alternation of languages were construed as cross-linguistic influence, one would expect bilingual children to use the gender of the translation equivalent noun when marking the determiner, but they do not.

In another recent study on mixed DPs, Liceras et al. (Reference Liceras, Spradlin and Fernández Fuertes2005) assume that the language of the functional category D will be the dominant one, in accordance with the grammatical features spell-out hypothesis formulated in Liceras (Reference Liceras2002) and Spradlin et al. (Reference Spradlin, Liceras and Fernández-Fuertes2003). The idea is that when bilingual children code-switch, the language with the richest array of uninterpretable (i.e. purely formal) features (Chomsky 1995, Reference Chomsky2001a,Reference Chomsky and Kenstowiczb) is the dominant one and therefore provides the surface realization of the functional category. Liceras et al. analyze data from Spanish–English bilingual children. Given that in Spanish the determiner carries two uninterpretable features, namely gender and number, and that the English determiner only carries the feature for number, in the mixed DP the determiner will come from Spanish, whereas the noun will be provided by English. This prediction is corroborated by the data: the sequence Spanish DET – English NOUN occurs in 98% of the cases. The authors also predict that if both languages have rich arrays, then no sequence will be preferred.

12.4 A radically monolingual approach to code-switching

12.4.1 Recent syntactic-theoretical approaches

A recent syntactic-theoretical approach to adult CS is offered by MacSwan (Reference MacSwan2000, this volume), in the Minimalist framework to syntactic theory (Chomsky 1995). He argues that CS can be explained by the same grammatical machinery that also underlies monolingual speech. MacSwan claims that the computational system and its operations are shared by both languages in the bilingual, whereas the lexicon is language-specific and thus has to be duplicated. The architecture of the bilingual speaker’s language faculty is represented in Figure 12.2.

Figure 12.2 The architecture of the bilingual (Italian–German) language faculty, following MacSwan (Reference MacSwan2000)

MacSwan suggests that there is a general ban on CS within a morphological word, which is captured by the PF Disjunction Theorem (“code switching within a PF component is not possible”; MacSwan Reference MacSwan2000:45). Morphology is located in the PF component of the grammar and the PF component is not a possible locus for CS since phonological rules have different orderings across languages. All cases of mixing within words are violations of the PF Disjunction Theorem and are therefore analyzed as cases of borrowing.

Puzzling for the PF Disjunction Theorem is the data presented by Cantone (Reference Cantone2007) who reports word-internal “switches” in child bilingual speech. Cantone observes that Italian nouns that are used in German utterances are adapted to German syllable structure. That is, the children elide the Italian vocal nominal ending: león is the noun used for adult Italian leone (“lion”), pappagall for pappagallo (“parrot”), scimm for scimmia (“monkey”), conchil for conchiglia (“shell-fish”), volp for volpe (“fox”), cavall for cavallo (“horse”), farfall for farfalla (“butterfly”), ranocc for ranocchio (“frog”). In the Italian context, the children sometimes add the vocal nominal ending to German nouns: Krona for adult German Krone (corona, “crown”). The age of the children ranges from three to four years, demonstrating that even older children exhibit this kind of word-internal creativity. Of course, one may interpret these instances as borrowing. Notice, however, that the lexical stem of the nouns is not phonologically integrated into German or Italian.

Also problematic for the PF Disjunction Theorem are the following instances of switches between the two languages, where genuine morphological processes are involved, like derivation and inflection and compounding. The German–Italian examples in (14) are taken from Cantone (Reference Cantone2007), the German–French examples from Veh (Reference Veh1990), the English–Spanish examples from Liceras et al. (Reference Liceras, Spradlin and Fernández Fuertes2005).

(14)

(14)

1. (a.)

   switch between French derivational prefix and German inflected verb

   deddy resucht

   Teddy again-look-for

   (Ivar, 2;4,9, Veh Reference Veh1990:98)

2. (b.)

   switch between German derivational prefix and French inflected verb

   dies on peut anmis- mise

   This one can on-put_[masc]-put_[fem]

   (Ivar 2;5,7, Veh Reference Veh1990:98)

3. (c.)

   switch between German noun and Italian diminutive suffix

   topfino

   little pot

   (Carlotta, 3;7,13, Cantone Reference Cantone2007:181)

4. (d.)

   switch between English adjectival stem and Spanish diminutive suffix

   sillito

   little silly

   (Simon, 3;10, Liceras et al. Reference Liceras, Spradlin and Fernández Fuertes2005:248)

5. (e.)

   switch between Italian nominal stem and German plural affix

   gelaten

   ice-creams

   (Aurelio, 3;0,19, Cantone Reference Cantone2007:181)

6. (f.)

   switch between a French noun and a German genitive suffix

   ça c’est Daniels et puis moi

   This it is Daniel’s and then me

   (Annika, 3,7,2, Veh Reference Veh1990:99)

7. (g.)

   switch between Italian verbal stem and German inflectional suffix

   e poi è finiert

   and then is finished

   (Aurelio, 4;0,28, Cantone Reference Cantone2007:181)

8. (h.)

   switch between German verbal stem and French inflectional suffix

   il a reité ne?

   he has ridden right?

   (Ivar, 2;0,29, Veh Reference Veh1990:98)

9. (i.)

   switch between Spanish verbal stem and English suffix

   I am laving myself

   I am washing myself

   (Leo, 3;3, Liceras et al. Reference Liceras, Spradlin and Fernández Fuertes2005:248)

There are other cases of word-internal mixes that are problematic. For example, compounding differs between Germanic languages and Romance languages. Whereas German exhibits morphological heads on the right edge of the complex word, French/Italian compounds are left-headed. Examples in (15) show compound-internal switches:

(15)

(15)

1. (a.)

   fruchtgust

   fruit-flavor

   cf. Italian gusto di frutta, German fruchtgeschmack

   (Carlotta, 4;3,23, Cantone Reference Cantone2007:181)

2. (b.)

   himbeerodore

   raspberry-scent

   cf. Italian odore di lampone, German himbeergeruch

   (Carlotta, 4;9,1, Cantone Reference Cantone2007:181)

3. (c.)

   kartoffelsoupe

   potato-soup

   cf. French soupe de pommes de terre, German kartoffelsuppe

   (Ivar, 3;5,28, Veh Reference Veh1990:99)

4. (d.)

   ein monsieurhut

   a man-hat

   cf. German herrenhut, French chapeau de monsieur

   (Ivar, 4;8,17, Veh Reference Veh1990:99)

5. (e.)

   ein omamasque

   a grandma-mask

   cf. French masque de grand-mère, German omamaske

   (Ivar, 4;11,14, Veh Reference Veh1990:99)

According to MacSwan’s model, the head of the compound should determine the position of the other elements. This is neither the case for kartoffelsoupe, nor for himbeerodore. Hence, these cases of CS seem to violate the language-specific requirements of the two languages involved. Nevertheless, they cannot be considered borrowings, because the children do not integrate the mixed words phonologically into the recipient language.

12.5 Conclusions

The present chapter has demonstrated that what researchers have criticized when analyzing adult CS is mirrored in the study of bilingual children’s language mixing. The investigation of children’s language mixing has taken a radically monolingual perspective, which implies that mixed utterances may be analyzed with the same linguistic machinery as monolingual utterances. The view of adult CS as a speech style which is constrained by nothing apart from the two grammatical systems involved has paved the way for monolingual analyses of children’s language mixing. Future research will have to take into account the difference between functional and lexical categories in child CS with the goal of distinguishing competence-driven from performance-driven explanations of the mixing datas.

13.1 Introduction

Code-switching (hereafter CS) can occur when signers of two sign languages interact. This is not surprising since CS is presumably a phenomenon that occurs regardless of the modality in which language is produced and perceived. Even so, the signed language researcher of CS is faced with challenges that may be unique to that modality. In particular, the question of how to attribute various signs or meaningful elements within an utterance (e.g. from Language A, Language B, both languages, or neither language) is among the main concerns. Admittedly, this phenomenon is not unique to signed language CS research, as evidenced by the discussion of congruence in spoken language literature and its role in CS (see Sebba, this volume). However, the potential for similarities between sign languages perhaps makes this issue much more pronounced in CS between sign languages.

Some signed languages are related historically, and this can be noted, in some cases, by examining lexical and grammatical similarities between the languages. However, regardless of the history of any combination of sign languages, there seem to be similar ways that signers use their bodies – not simply their hands – to create meaning across such languages, and this results in the production – within a signed stream – of elements whose meanings are relatively transparent to an interlocutor. Essentially, such characteristically transparent communicative devices exist across sign languages, and in some cases they take on linguistic roles. For example, an extended index finger directed at the signer herself, often at the chest but possibly at the face, usually acts as the first person singular pronoun, and points in the signing space often indicate locative references such as here or there. So-called “classifier” constructions and bodily actions that appear mimetic in nature are also found in the utterances of signers of different sign languages, and those constructions and actions are often difficult to attribute to one particular sign language as opposed to another. If one also considers the articulation, within the sign stream, of common gestures that are used throughout various cultures (e.g. the thumbs-up gesture to indicate that something is good), the degree to which meaning creation is transparent across sign languages – even those that are unrelated – is significant. CS researchers who work on signed language data must carefully consider a broad spectrum of meaningful devices that signers produce because they influence the ways in which CS analyses are performed. Since users of spoken languages can also accompany their speech with points and gestures, it would appear that such productions could also present a challenge for researchers of CS in speech – not only for sign linguists. In essence, the signed modality forces us to consider ways in which linguistic and gestural devices interact, and this could be extremely valuable to CS analyses of spoken or signed languages.

In addition to communicative devices that are somewhat similar across sign languages, it appears that various linguistic structures of signed languages are more similar to each other than is the case for spoken languages, and this holds true even when one considers unrelated sign languages. One could suggest that this is the case at the level of phonology, morphology, and even syntax. As a result, whether or not a particular form can be described as a code-switch could be questionable. This situation may be akin to types of CS that occur between two historically related spoken languages like Spanish and Portuguese, but it is perhaps very different for examples of CS between structurally diverse languages.

There is at least one other major challenge that is faced by the researcher of signed language CS, and it also relates to the primary question of how to determine what language is operating at any one time during the articulation of various elements within the sign stream. This challenge stems from the fact that the articulators that a signer uses – body parts that allow sign languages to express meaning in certain ways – differ from those used in spoken languages. In essence, a signer can use more than one body part (e.g. hands, arms, head, torso) simultaneously to create meaning, and this fact influences how CS is examined in signed language research.

13.2 Notions to consider: differences between sign and speech

Despite similarities in various facets of linguistic structure between signed and spoken languages (e.g. the existence of phonological primitives, various word-formation processes, and syntactic structures), there are some noteworthy differences between languages across the two modalities. Several of these likely stem from having the hands, arms, and other upper body parts as articulators as well as from the use of the immediate area in front of the signer as an important space in which signs are articulated.

The simultaneous nature of signed language has been recognized since the beginning of linguistic research on American Sign Language (ASL) and other sign languages (Fischer Reference Fischer, Rohrer and Ruwet1974; Klima and Bellugi Reference Klima and Bellugi1979). For example, morphemes that communicate person, number, and aspectual information can occur concomitantly with some verbs, and a signer’s two hands can be used to simultaneously articulate two different classifier constructions, referred to hereafter as polycomponential signs, following Schembri (Reference Schembri and Emmorey2003) and Slobin et al. (Reference Slobin, Hoiting, Kuntze, Lindert, Weinberg, Pyers, Anthony, Biederman, Thumann and Emmorey2003). A signer can also produce non-manual signals (e.g. mouth and head movements, torso shifts, and patterns of eyegaze) simultaneously with a lexical sign in order to modify that sign (or phrase). As an example, a non-manual signal such as an adverbial mouth movement can co-occur with verb signs. The mouth and lips can also serve to articulate, without voice, a spoken language word while the signer produces a semantically equivalent sign, and this is commonly referred to as “mouthing.” Even if one considers tonal contrasts in some of the world’s languages and prosodic features that provide meaningful information, spoken languages do not tend toward exhibiting simultaneity to the degree that signed languages do.

One reason for differences between signed and spoken languages may lie in the purported speed of sign production versus spoken word production. Klima and Bellugi (Reference Klima and Bellugi1979) claim that, on average, a spoken word can be uttered in half the time required to articulate a sign.¹ Meier (Reference Meier, Meier, Cormier and Quinto-Pozos2002) hypothesizes that the rate of signing versus speaking plays a prominent role in the simultaneous nature of signed languages because it discourages sequential affixation.² Essentially, an “articulatory constraint may push natural sign languages, such as ASL, in a particular typological direction, that is, toward nonconcatenative morphology” (Meier Reference Meier, Meier, Cormier and Quinto-Pozos2002:8).

Another factor that may play an important role in leading to the simultaneous nature of signed languages has been described in terms of the amount of information that can be communicated simultaneously in one modality versus the other. Meier (Reference Meier, Meier, Cormier and Quinto-Pozos2002:10) suggests that, “at any instant in time more information is available to the eye than the ear, although in both modalities only a fraction of that information is linguistically relevant.” Emmorey (Reference Emmorey2002) is consistent with Meier’s argument in her claim that it is easier to visually perceive spatially disparate information in parallel than to perceive and decipher different types of auditory information simultaneously. In other words, it is easier to perceive complex visual displays at once than auditory signals that may contain disparate types of information.

The use of three-dimensional space in the articulation of sign seems to also lead to some interesting differences between sign and speech. Sign languages allow for the simultaneous communication of various types of information about one or more objects. As mentioned earlier, a signer can articulate a polycomponential sign with one hand and a different one with the other hand, and the two hands interact in specific ways (Supalla Reference Supalla and Craig1986).³ Such articulations, referred to here as entity polycomponential signs, can provide information about motion and/or location of the objects, including to what type or class each item belongs. The kinds of productions that have been labeled polycomponential signs are also used to describe how objects are handled (handle polycomponential signs) as well as how objects can be described in visual–geometric ways (size and shape specifiers) (Emmorey Reference Emmorey2002; Schembri Reference Schembri and Emmorey2003).

In addition to the use of 3-D space by signers when they articulate polycomponential signs, users of all sign languages also have access to the gestural medium for meaning generation. As a result, signers can alternate linguistic signs with non-linguistic gestures, and the signs and gestural material can also co-occur in some cases – such as with deictic pointing and verbs that indicate person and number, referred to commonly as “agreement” verbs or “indicating” verbs. The gestures themselves are sometimes culturally specific emblems that are also produced by members of the hearing community (McNeill Reference McNeill1992). However, the gestures can also be deictic and pantomimic in nature. The latter are particularly intriguing because they tend to be used regularly in sign languages. Signers across different sign languages produce similar mimetic gestures that alternate with linguistic material, which are referred to by some researchers as constructed action. Constructed action has been described for sign as the way in which a signer uses her body to depict aspects of an animate entity (Metzger Reference Metzger and Lucas1995; Aarons and Morgan Reference Aarons and Ruth2003). For example, a signer might “act” like another person or an animal when describing something about that being or something that occurred. Clark and Gerrig (Reference Clark and Gerrig1990) describe a similar phenomenon as an accompaniment to spoken language use, and they develop an argument for why demonstrations, as they call these mimetic actions, function as quotations. The alternation of gestural material such as emblems and constructed action with linguistic/grammatical material might be rule-governed, although such systematic relationships have been addressed only minimally in the literature (e.g. Aarons and Morgan Reference Aarons and Ruth2003). The fact that signers have access to gestural resources within the same channel of communication poses a challenge for the researcher who is analyzing CS data, as will be demonstrated later in this chapter.

With regard to linguistic structure, some authors have suggested that sign language phonologies are more similar to each other than spoken language phonologies when compared cross-linguistically. Lucas and Valli (Reference Lucas and Valli1992) note that signs referencing names of foreign countries have become incorporated into ASL from other signed languages, but the phonologies of the source languages are so similar to the phonology of ASL that it is difficult to determine if the incorporation should be considered a lexical borrowing or an example of CS. Borrowings in spoken language have often been characterized by the phonological integration of the borrowed word into the phonology of the other language, but this integration may not be so evident in signed language. For instance, the sign italy as signed in Italian Sign Language (LIS) has now been incorporated into ASL, in some cases replacing the older sign for Italy. Lucas and Valli suggest that LIS and ASL have similar phonological inventories of handshape, palm orientation, and location, and this is true even though they are not related or mutually intelligible as languages. They also note that the languages may have similar segmental structure. One part of the authors’ rationale for claiming that sign languages have similar phonologies lies in the assertion that such languages have many more basic components (i.e. basic handshapes, movements, places of articulation, etc.) than the sets of inventories that spoken languages contain. They cite as evidence the suggestion of a colleague (Robert Johnson) that:

The paradox is that while there may be more basic components in signed language, various signed languages in their present forms seem to share a significant percentage of those large sets. Visual iconicity in signed languages – such as the use of deictic forms, polycomponential signs, and construction action – perhaps contributes to this situation.

In terms of grammatical items, there are cross-linguistic differences in some aspects of sign language morphology and syntax, such as word order (Newport and Supalla Reference Newport, Ted, Emmorey and Lane2000), the existence of auxiliary verbs that use locations in the signing space to indicate subjects and objects (Rathmann Reference Rathmann2000; Quadros Reference Quadros1999), and even the grammar of negation (Pfau Reference Pfau, Meier, Cormier and Quinto-Pozos2002). By some accounts, syntax is the level of structure in which sign languages may most closely resemble spoken languages. Yet, sign languages seem to demonstrate similar morpho-syntactic structures. They all appear to have different categories of verbs, e.g. verbs that indicate the subject and object of the verb by movement through space and verbs that rely on word order for the assignment of case. Other examples of cross-linguistic similarities concern aspectual modifications to verbs, the use of pronouns, and the use of polycomponential signs.

Additionally, sign languages appear to possess a base level of lexical similarity that is greater than that found for spoken languages, a fact that is likely influenced by a significant degree of iconicity in sign languages. Iconicity is a complex phenomenon, but for the purposes of this chapter it can be defined as the ways in which a signer creates visual correspondences between her own body (hands, arms, torso, head, etc.) and the referent. The modest degree of lexical similarity between sign languages is even true for sign languages with no known historical or genetic relationship. The potential for visual iconicity in the signed modality influences signed languages in this regard. However, there are many lexical items in sign languages that are not considered iconic and others that have become less iconic over time. A higher degree of iconicity can make it difficult for the CS analyst to determine if a particular sign – especially if it is very iconic – is really a sign of one language and not the other or just a visually meaningful way of representing a concept that may not be a lexical item in either language. While common methods for determining lexical similarity across sign languages are useful, they are also somewhat arbitrary and may not reflect the ways in which signers recognize and process signs. For instance, similarity has been determined by comparing articulations across the parameters of sign formation, and a similarly articulated sign is one that is determined to share at least two of the three values of the major phonological parameters (handshape, movement, and place of articulation) (Guerra Currie et al. Reference Guerra, Anne-Marie, Meier, Walters, Meier, Cormier and Quinto-Pozos2002). As will be noted later, this method, while useful for various analyses, may allow for important information to be overlooked.

In summary, when one considers the various ways in which sign and spoken languages differ from each other, it becomes clear that analyses of CS in signed language are challenging. There are times when it is not clear how to differentiate the languages used in a particular utterance. However, there are also instances when signs that are unique to one or the other language can be identified, and sometimes a switch occurs at a location where such signs occur in sequence. This is, perhaps, the best place to begin a discussion of CS in sign. This chapter provides some illustrations of what seem to be clear cases of CS between two sign languages, although the presentation will also include various examples of issues in the labeling of meaningful elements.

13.3 Code-switching in sign

13.3.2 CS between signed languages

Thus far, all the examples of CS that have been described concern the mixing of a signed language and some form of a spoken language. It seems that little work has been done on the mixing of two signed languages, and examples of sign–sign CS are mostly lacking in the literature. One work that does provide some examples of such phenomena is Quinto-Pozos (Reference Quinto-Pozos2002), and it focuses on contact between ASL and Mexican Sign Language (LSM) along two areas of the Mexico–US border in Texas. There are other areas of the world where one would expect contact between two signed languages, although it appears that no published works exist that document such contact. One such area might be along the border of two provinces of Canada, Quebec and Ontario, where different signed languages, Quebec Sign Language (LSQ) and American Sign Language (ASL), are used. Contact between signed languages may also occur in parts of Spain, where Spanish Sign Language (LSE) and Catalan Sign Language (LSC) are used by populations of Deaf signers.

For the study of LSM and ASL contact, Quinto-Pozos (Reference Quinto-Pozos2002) videotaped interactions between users of Mexican Sign Language (LSM) and American Sign Language (ASL) who live on the United States side of the US–Mexico border. Both LSM and ASL have been reported to be historically related to the Old French Sign Language (OLSF) of the 1800s (Guerra Currie Reference Guerra and Anne-Marie1999; Adams Reference Adams2003), although the two languages are distinct and not mutually intelligible (Faurot et al. Reference Faurot, Dellinger, Eatough and Parkhurst1999). Yet, there do exist lexical and grammatical similarities between the two languages.

The CS data reported in Quinto-Pozos (Reference Quinto-Pozos2002) come from deaf signers who were fluent bilinguals in the two languages and others who were mostly proficient in one of the two languages. The data collection involved group discussions (four participants per group in each of two locations) and one-on-one interviews, and those sessions were examined for various contact phenomena in the signed modality.

13.3.2.1 Reiterative CS

One type of CS described in Quinto-Pozos (Reference Quinto-Pozos2002) is the switching of synonymous signs. In these cases, each of these code-switched elements was produced after a participant would articulate a semantically equivalent sign from the other language that differed in form. Of the 40 switches of this type from 64 minutes of conversation, more than half (n = 23) were nouns, one-fifth (n = 8) were verbs, one-eighth (5) were adjectives, and there were also a couple of possessive pronouns and adverbs. A seemingly similar type of CS in spoken language contact situations has been termed reiteration (Auer Reference Auer, Milroy and Muysken1995; Eldridge Reference Eldridge1996; Pakir Reference Pakir1989; Tay Reference Tay1989). This is the phenomenon of a message in one code being repeated in another code. Various social functions have been attributed to the phenomenon of reiterative CS and among them are: negotiation of a collective social identity, accommodation, amplification of a message, emphasis, reinforcement or clarification of a message, and attention-getting, as in the regulation of turn-taking (Pakir Reference Pakir1989; Tay Reference Tay1989; Auer Reference Auer, Milroy and Muysken1995; Eldridge Reference Eldridge1996). In the LSM–ASL data that were analyzed by Quinto-Pozos (Reference Quinto-Pozos2002), the CS seems to have served several of the social functions just mentioned such as emphasis, clarification, accommodation, and reinforcement. However, there are also cases where the functions of switching are not clear.

In many cases, CS (not only reiterative switching) can serve what Appel and Muysken (Reference Appel and Muysken1987:119) call a directive function – the desire to “include a person more by using her or his language.” This directive function that Appel and Muysken describe is not unlike the concept of accommodation that Pakir (Reference Pakir1989) described as a function of reiterative CS. The first example from Quinto-Pozos (Reference Quinto-Pozos2002) can be seen in (4), where the code-switched sign is bolded. In these examples, LSM signs are represented by Spanish words in capital letters and those from ASL are indicated via English. Points are indicated by their form (e.g. “point to finger”) or by their function in the case of pronouns (e.g. ME/YO).

(4)

(4)

point: middle finger (for listing) TOMATO TOMATE ADD-INGREDIENTS MIX gesture: “thumbs-up”

“(. . . and then you take) tomatoes and you add them to the other ingredients and mix everything together. It’s great.”

Example (4) contains a few items that do not allow for easy classification as elements from LSM, ASL, both languages, or neither language. One example is the point that begins the segment; it seems to be a common listing strategy that is not attributable to only one of the sign languages. Also noteworthy is that the ASL signs indicated as add-ingredients and mix are quite transparent (or iconic), although they have been labeled as ASL simply because they were not confirmed by the author to also be LSM signs. It is likely that both of those purported ASL signs would be understood by signers of both languages. Example (4) does contain CS, however, and that is the focus of the following discussion.

In (4), the bilingual interviewer was mostly looking in the direction of two users of LSM who were raised in Mexico as they recapped cooking instructions that were presented earlier by another participant. The interlocutors who engaged the signer frequently produced LSM signs in other segments of the discussion, which is why the signer may have made a conscious decision to add the LSM nominal sign tomate after the ASL sign tomato. There was a very brief pause between the sign tomato and tomate, which gives the code-switched item a certain degree of emphasis. In some respects, the code-switched sign could also be viewed as a clarification – a sign used to clarify an ASL sign that might not be entirely familiar to at least one of the other participants. Also, note that the final meaningful element in (4) is the emblematic gesture “thumbs-up,” which provides a positive comment about what had just been described. Whether or not such an element should be considered an LSM or ASL sign – having become lexicalized into either or both of those languages – is another question that should be addressed, and this would also apply to other emblems that are used within the sign stream.

In another example of CS from the group discussions, the interviewer code-switched a verb while asking a question about what one of the participants regularly does for her birthday. The sequence of signs that contains that verb appears to be a serial verb construction, a type of syntactic construction that is common in ASL and perhaps other sign languages (see Supalla Reference Supalla, Fischer and Siple1990). The example can be found in (5), and the code-switched item is in bold. Like (4), the example given in (5) shows the use of deictic points to a second-person singular interlocutor. Such points would be produced in either language, although they are also common outside of the two languages within the gestural communication of hearing people. They have been labeled in (5) as pronouns from both languages.

(5)

(5)

TÚ/YOU CUMPLEAÑOS TÚ/YOU HACER FORM-GROUP INVITE 2sg birthday 2sg do INVITAR SELF TÚ/YOU INVITAR invite yourself 2sg invite

“For your birthday, do you usually invite people to get together? Do you do that (yourself)?”

In (5), there are several clear switches from unique signs in one language to unique signs in the other. For instance, hacer (“do”) to form-group is the first clear switch, and invitar to self is the second. It very well could be the case that the second-person singular switches were influenced by only one of the sign languages, although the surface forms do not allow for such a determination.

In (5) the signer did not pause, even briefly, before the code-switched item. Thus, this example does not exhibit the emphasis that characterized the CS in (4). Yet, this example might still function as accommodation or even identification with the other signer. The interlocutor who held the signer’s gaze during this sequence was one of the participants who produced the most LSM in the border data collection sessions. Further, the interview session with that interlocutor was characterized by relatively large amounts of LSM production. As in that interview session, the interviewer, during this example from the group discussion, may have presumed that this particular interlocutor preferred LSM and thus made an effort to produce LSM signs. This type of CS can also be described as serving a reinforcing function, which is one of the roles that reiterative CS has been claimed to perform.

In example (5) it is not clear what function the code-switched element served. This is also true of other examples of reiterative CS that occurred in the group discussions and interviews as reported in Quinto-Pozos (Reference Quinto-Pozos2002). During a discussion of whether or not participants’ families are Deaf or hearing and how the participants communicate with their families, a Deaf female participant who was raised in Mexico commented on the fact that most of her family are hearing. The example is given in (6), and the code-switched item, the LSM noun familia, is in bold. Similarly articulated signs (those that differ by the value of one phonological parameter at most but that are similar in meaning) in LSM and ASL are represented in capital letters in Spanish and English separated by a slash.

(6)

(6)

NO/NO ME/YO NO/NO++ ME/YO gesture: shake-finger DEAF/ SORDO __ head shake for negation __ gesture: “wave hand to negate” ME/YO FAMILY FAMILIA MY/MI gesture/emblem: “well”

“As for me, my family is not Deaf. Oh well.”

In (6), there was no pause between family and familia. The sign familia was not stressed and no other means were used to draw attention to this sign. This does not seem to be a clear case of emphasis. Further, while the female participant signed familia she was looking at another participant who signed mostly ASL during the group discussion and interviews. Thus, this code-switch does not seem to be a case of accommodation either. Perhaps this instance of CS was intended to display an identification with the interlocutor, but there are no explicit features (emphasis of the sign, a pause, change of eyegaze, etc.) that would suggest what the signer’s intent was when she produced this code-switch. As can be seen, the reason(s) for using CS are not always clear. Sometimes there are no explicit features that would suggest that the code-switch was deliberately produced for a specific reason(s). Thus, lists of CS functions do not seem to account for all instances of CS.

Another point that is noteworthy about (6), in consideration of the challenge for determining how to attribute various elements in the sign stream, is that the only signs that are unique to the two languages are the signs family and familia – the actual location of the CS. So, while it is possible to note that the signer code-switched here, it is not possible to determine how to label the other meaningful elements (points, similarly articulated signs, and common gestures) within the sequence. This is a problem for CS analyses that rely on clearly identifying the source language for each lexical item.

13.3.2.2 Non-reiterative code-switching

The corpus of US–Mexico border data collected by Quinto-Pozos also contains examples of CS that are not of the reiterative variety. These are presented here to further illustrate why it is often difficult to clearly attribute a meaningful element from the sign stream for purposes of CS analyses. In the examples presented in this section, the code-switched item does not follow a semantically-equivalent sign from the other language, as was done in examples (4) through (6). In some cases the switches are of single signs, but the switch might also contain a sequence of signs and/or other meaningful productions (e.g. polycomponential signs, gestural productions, and/or constructed action).

In the first example of non-reiterative CS, the participant describes how candy from Mexico is quite different from that of the United States. The signer is left handed, and that is her default dominant hand for signed language production. But, as will be noted in (7), she switches hand dominance for a short sequence of signs for comparison purposes, which is a common non-manual strategy for providing comparisons in ASL and perhaps in other sign languages as well. In (7) through (9), the bolded item represents a clear switch from a previous sign or sequence in the other language. Any sign unique to either LSM or ASL that immediately follows a similarly articulated sign is not represented in bold in (7) because it is not clear if the similarly articulated sign should be labeled an LSM or ASL production using the current methods of sign analysis. Other transcription conventions pertinent to this example include: [lh] or [rh] to indicate an articulation with the signer’s left or right hand, and the use of “+” to indicate a single repetition of the sign, “PS2” (or Polycomponential Sign 2) to indicate a possible item from the set of so-called handle classifier forms, and “CA:” with a brief description of the signer’s enactment to indicate the use of constructed action.

(7)

(7)

BUT [lh]: point: upward and leftward FOOD/COMIDA DULCE there sweet/candy DIFFERENT/DIFERENTE HOT+ PICANTE-CHILE DELICIOSO spicy delicious [rh]: point: downward CA: signer tasting candy/PS2: holding a small item here LOUSY BUT [lh]: CHOCOLATE [ASL] DELICIOUS point: upward and leftward CHOCOLATE [ASL] LITE HIGH DULCE RIGHT point-TV3 HIGH sweet/candy

“However, the candies in Mexico are different; they are spicy and delicious. Here, they are lousy. But, the chocolate here is also sweet. There the chocolate is lighter and not so full of sugar as it is here.”

As can be noted, (7) contains two examples of LSM-ASL similarly articulated signs (food/comida and different/diferente), and it also contains several examples of pointing, but there are also several unique signs from each of the languages. LSM signs include dulce, picante-chile, and deliciosa. ASL signs include but, hot, lousy, chocolate, lite, right, and high. The passage begins with an ASL sign as a conjunction, and ASL signs outnumber LSM signs. One could claim that the passage seems to contain more ASL than LSM (lexically and in terms of grammatical function words), although the similarly articulated signs, points, and brief use of constructed action present challenges for CS analyses.

The next two examples include instances of polycomponential signs. The sequence in (8) is about the preparation of a food dish in Mexico that does not need to be cooked before serving; [bh] indicates the signer’s articulation with both hands; PS3 is a size and shape specifier.

(8)

(8)

THAT (response to the interviewer) [bh]: PS3: meat exiting a grinder [bh]: PS3: motion of gears that are grinding something TOMATO ONION⁴ PICANTE/CHILE MIX spicy [rh]: PS3-shape of bowl,[lh]: PS2: stirring as if with spoon NOT NEED FUEGO/FIRE NOT NEED FUEGO/FIRE COLD/FRÍO EAT/COMER DELICIOSO delicious

“That’s it. It’s the tomato, onion, chile that is mixed together and you stir it up in a bowl. You don’t need to cook it; you just eat it cold. It is delicious.”

Note that the sequence in (8) begins with an ASL sign (that), which is an affirmation of what the interviewer had just signed. The signer goes on to describe a grinding action with polycomponential signs and follows that with two ASL nouns and an LSM noun. It is not clear if the polycomponential signs should be analyzed as LSM structures, ASL structures, or both. Three other elements surface that are difficult to label as either ASL or LSM, and they are the highly iconic sign mix, a polycomponential sign depicting the side of a large bowl, and the signer showing the mixing of something in the bowl. The final sign of the entire sequence is the LSM sign delicioso, an adjective that describes the food that is prepared in that manner. In some respects, this sequence appears to have more of an ASL character because of the several ASL noun signs (tomato, onion), and the negation and modal signs (not need) that were signed twice. However, it should be noted that even though fuego/fire is coded as being a similarly-articulated sign, there are nonetheless some differences between fuego and fire, specifically, articulations for hand internal movement, path movement, and whether or not the fingers are fully extended or bent. See Table 13.1 for a comparison of the two signs. Even though fuego/fire seem to differ in several ways, an analysis that would consider only the three major parameters of sign formation for the determination of similarly articulated signs might lose the distinctions between the signs (either because the two handshapes would be considered variants of a 5-handshape or because the up-and-down path movements would be considered similar, even though one occurs with a circular movement and one does not). However, if one were to consider more fine-grained phonetic analyses of the two signs, the results would likely suggest that the sequence with the purported similarly articulated sign should probably be shown as in (9):

(9)

NOT NEED FUEGO NOT NEED FUEGO

Then, the code-switching analysis could focus on the switch between a verb (need) and its object (fuego).

Table 13.1 Differences between LSM FUEGO and ASL FIRE

The final example, given in (10), also includes polycomponential signs. This sequence describes a participant explaining to the interviewer and the others in the group that it is easier to understand a written Spanish recipe than a written English recipe. In this example, the item denoted as “PS1” should be considered within the entity category of polycomponential signs.

(10)

EASY (10+)

BUT [lh]: ME/YO MÉXICO THINK MÉXICO FÁCIL

[rh]: point-downward TOUGH INGLÉS PS1:flat object CA:signer looks at paper

[lh]: UNDERSTAND EXPLAIN LONG/LARGO PS1: paper CA: signer looks at paper

WRONG TIME TWO/DOS WRONG THREE TIME BIEN/GOOD

“That’s easy [referring to reading Spanish]. I think it’s easy in Mexico. Here [United States] English is tough. In order to understand something written in English, it takes a long time. Sometimes I get something wrong two times, then the third time is fine.”

In (10), the signer begins with ASL signs, interjects LSM signs for the country Mexico, but then also uses a mixture of LSM and ASL signs in the next few signs. The adjective fácil (“easy”) is in a clause (assumedly beginning with the conjunction but and ending with the adjective) that has only two signs that are clearly from LSM: méxico and fácil. Although, as mentioned earlier, the country signs used in foreign sign languages (e.g. japan in Japanese Sign Language, méxico in LSM, etc.) are more common in ASL than they used to be; perhaps they could currently be considered borrowings. This complicates matters because now it is not clear if the clause is mostly LSM or ASL, and there are no function word signs in this sequence to provide information about which grammar is being utilized at various points in the sign stream. Further, it is not clear how to label the source language of the polycomponential sign (to represent a paper or other written document) and the constructed action of the signer gazing at the paper.

Based on the data presentation and the brief discussion of examples (7), (8), and (9), it is clear that some examples of CS are quite challenging to analyze because of the issues raised earlier. Note that there is little discussion of non-manual signals (mouth gestures, eyebrow movements, torso shift, etc.) in these passages, which would present yet another example of simultaneous articulations that would need to be examined. Additionally, possible switches could be lost because of the current system of classification for similarly articulated versus non-similarly articulated signs. Also, the segments that contain polycomponential signs and constructed action are particularly difficult to attribute to one language or the other. All of these issues create challenges for CS analyses in sign.

13.4 Conclusion

As was suggested throughout the data presentation section, CS analysts of sign data are faced with the challenge of determining how to label some of the meaningful elements from a signed conversation. And, such labeling should occur before the data can be examined within any particular model or theory. Current frameworks for CS analyses are primarily based on sequential analyses of meaningful elements (i.e. words, bound morphemes, etc.) without taking into account the alternations with gestural material or influences from visual iconicity that occur in the signed modality. There are also challenges in sign analyses that have to do with simultaneity in that modality. These issues pose challenges and difficulties for analyses of sign data.

Yet, in some cases, sequential CS can be identified and analyzed for sign. This is true for the LSM–ASL reiterative switches presented in examples (4), (5), and (6), and these examples do not seem to contain simultaneous articulations that would leave the researcher wondering how to label each of the code-switched signs. They also contain sign pairs – the reiterative switch and the sign that precedes it – that are clearly articulated differently in the two languages. Even though ASL and LSM are related historically and have similar phonologies, the lexical items in those examples differ from each other, which allows the analyst to determine when the signer is producing one language versus the other. However, those examples also contain the use of gesture (both in the form of widely used emblems and also in the form of constructed action) that alternates with signs, and some of the signs are highly iconic. This can be problematic for language labeling.

Signed languages have some structures that pose challenges for the CS researcher. In particular, the simultaneous nature of sign (e.g. polycomponential signs, code-mixes, and code-blends), the apparent similarity of some sign language structures, and the interaction of signs with non-linguistic gestures need to be considered carefully. One way to address those challenges is to produce more fine-grained descriptions of the phonetic, phonological, morphological, and syntactic structures of signed languages. Knowing, in specific ways, how sign languages differ from each other will allow for the examination of possible examples of CS between such languages. Further, lexical comparisons between sign languages need to concern themselves with more than the major parameters of sign formation; specific details of orientation, finger positions, contact locations, and the like are also necessary. More cross-linguistic work in sign could perhaps help to understand how polycomponential signs differ from each other (if they do), and offer suggestions about how researchers can identify differences between such signs in different sign languages. Research on how specific languages constrain or govern the use of constructed action as it interacts with the linguistic system is sorely needed. Work of this nature can inform CS analyses in sign, and will allow for the inclusion of sign data in theories and models of CS. Assumedly, theories of CS should be equally applicable to sign and speech data, but that remains to be confirmed with more empirical data.

Finally, it seems that all CS researchers should be faced with the challenge of accounting for multi-modal data. Two possible questions for such a line of inquiry could be: how does the use of gesture and demonstration influence the way people code-switch in their communication? And, how does the use of spoken words and phrases interact with emblematic gestures in spoken language conversations? Signed language data can provide exciting opportunities to consider how non-verbal ways of communicating interact with linguistic systems.