Bilingualism. Memory, and Emotion

by Hugh Knickerbocker and Jeanette Altarriba
University at Albany, State University of New York

Several models of bilingual memory describe the interplay between lexical and semantic stores of memory in bilingual individuals attempting to comprehend and produce speech. However, while these models have emphasized the pattern of connections between general linguistic and semantic clusters across languages, only a small amount of work has investigated the perception of emotion across languages. Numerous lines of research have showcased emotion effects and have provided insight into the effects of emotion and language on semantic and autobiographical memory.

Multiple studies have investigated the automatic activation of emotion words across first (L1) and second (L2) languages. The findings of these studies are heavily influenced by the pattern of language dominance exhibited by the sample of participants. Procedures such as the Stroop task, where participants report the presentation color of a word rather than the word itself (e.g., the word ‘fear’ presented in the color blue is responded to more slowly, as opposed to the word ‘box’ in the color blue), or priming, where response to a target word is faster when it is preceded by a related prime word (e.g., participants respond faster to ‘depressed’ when primed by ‘sad’) as compared to an unrelated word, have shown similar emotion effects when bilinguals are fluent and regularly use both languages (Fazio, Sanbonmatsu, Powell, & Kardes, 1986; Sutton, Altarriba, Gianico, & Basnight-Brown, 2007). However, emotion effects can be limited to one language if bilinguals have a clear dominant language.

Research into the automatic processing of emotional language has led to even more interesting insights into the structure of autobiographical memory in bilingual individuals. Research has utilized skin conductance response (SCR) where changes in the conductivity of the skin are measured as an indication of autonomic nervous system responses. Stressors should result in an autonomic nervous system response that can be detected by monitoring sweat levels on the skin. Increases in sweat increase the conductance across participants’ skin. Bilingual research using SCR measures provides insight into the automatic processing of emotion across L1 and L2. This research has found that late learners of L2 exhibited SCR effects to emotion words (e.g., sadness) in both their L2 and L1, but only exhibited SCR effects for reprimands (e.g., shame on you) when they were presented in their L1. During debriefing interviews, participants reported unique automatic memory retrievals from their youth when presented with reprimands. These memories were typically events where participants were reprimanded as children by a family member or other authority figure. These findings provided strong evidence of the existence of language-specific memories that can best be retrieved through the use of a specific language (Harris, 2004; Harris, Ayçiçeği, & Berko-Gleason, 2003).

Research into conscious retrieval and autobiographical memory has also provided evidence of language-specific memories. Investigations have found that memories tend to be more available for retrieval in the language in which they originally occurred. Bilinguals generally provide memories with a greater level of detail and elaboration when retrieving a memory in the language in which the event occurred. Studies of the autobiographical memory of bilingual immigrants who changed their daily language usage as a result of emigrating have bolstered this view. This research has examined the ‘reminiscence bump’ which is a time period between the approximate ages of 10 and 30 that results in a greater number of autobiographical memories. The reminiscence bump tended to shift to match the time period of emigration. More interesting, the memories of the participants were clearly divided between their two known languages. Memories before immigration (and the reminiscence bump) were stored in participants’ original language. Later memories that were from a post-immigration period were stored and more easily retrievable in the language that participants were forced to switch to as a result of their immigration event (Schrauf, & Rubin, 1998, 2000, 2001).

The processing of emotional speech by bilinguals has already begun to clarify some of the idiosyncrasies of semantic and autobiographical memory retrieval and structure. We now know that emotional stimuli can have the same impact regardless of language, as long as participants have similar levels of fluency and daily usage in all known languages. However, some emotional connections remain unique and highlight features of autobiographical memory.

Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986). On the automatic activation of attitudes. Journal of Personality and Social Psychology, 50, 229-238.

Harris, C. L. (2004). Bilingual speakers in the lab: Psychophysiological measures of emotional reactivity. Journal of Multilingual and Multicultural Development, 25, 223-247.

Harris, C. L., Ayçiçeği, A., & Berko-Gleason, J. B. (2003). Taboo words and reprimands elicit greater autonomic reactivity in a first language than in a second language. Applied Psycholinguistics, 24, 561-579.

Schrauf, R. W., & Rubin, D.C. (1998). Bilingual autobiographical memory in older adult immigrants: A test of cognitive explanations of the reminiscence bump and the linguistic encoding of memories. Journal of Memory and Language, 39, 437-457.

Schrauf, R. W., & Rubin, D.C. (2000). Internal languages of retrieval: The bilingual encoding of memories for the personal past. Memory & Cognition, 28, 616-623.

Schrauf, R. W., & Rubin, D.C. (2001). Effects of voluntary immigration on the distribution of autobiographical memory over the lifespan. Applied Cognitive Psychology, 15, S75-S88.

Sutton, T. M., Altarriba, J., Gianico, J. L., & Basnight-Brown, D. M. (2007). The automatic access of emotion: Emotional Stroop effects in Spanish-English bilingual speakers. Cognition and Emotion, 21, 1077-1090.


Find out more about Memory, Language and Bilingualism here.

Sapir, Whorf, and the hypothesis that wasn’t

written by Aneta Pavlenko, Temple University



One of the linchpins of human information-processing are the frames of expectation we apply to the constant flow of information. These frames allow us to impose meaning on the things we see, hear, or read and to position ourselves with regard to ideas and arguments. In the case of the Sapir-Whorf hypothesis (SWH), these frames require us to adopt one of the three recognizable positions: for (which may brand us as radicals), against (a marker of a skeptic or a rational thinker), or in-between (a sign of a temperate scholar willing to consider the pros and cons of everything). The adoption of conventional frames of expectation saves us a lot of valuable time – once we know (or think we know) what each position means and where each party stands, we can jump right in the middle of any argument and hammer in our own point of view. I have experienced the power of this conventionality firsthand when I gave an interview about my book, The Bilingual Mind, to François Grosjean at Psychology Today: some readers immediately branded it as defense of the relativist cause (at best) or ‘nonsensical relativism’ (at worst). Yet there is a downside to unquestioning adoption of conventional frames of expectation – it leaves us vulnerable and unprepared for changes in the terms of engagement. The Bilingual Mind is a case in point: I do not argue for or against the SWH because I do not see it as a legitimate scientific phenomenon. The purpose of the book is to show that the SWH ‘as we know it’ is a phantom, if not a fraud, and has little to do with questions that preoccupied Sapir and Whorf.

The manufacturing of consent on the SWH began when Sapir and Whorf passed away and their ideas landed in the hands of others. Driven by the desire to make complex notions, articulated by linguistic anthropologists, fit experimental paradigms in psychology, Roger Brown and Eric Lenneberg transformed them into two ‘testable’ hypotheses. These hypotheses – one ‘weak’, one ‘strong’ – appeared in their definitive form in Brown’s (1958) book Words and things:

linguistic relativity holds that where there are differences of language there will also be
differences of thought, that language and thought covary. Determinism goes beyond this
to require that the prior existence of some language pattern is either necessary or sufficient to produce some thought pattern. (p. 260)

Soon the newly-minted SWH took on a life of its own, multiplying and reproducing itself in a myriad of textbooks, articles, lectures, and popular media. Yet ideas don’t travel easily across disciplines and Brown’s and Lenneberg’s reformulations departed from Sapir’s and Whorf’s original arguments in several ways. To begin with, they articulated the hypothesis in monolingual terms (while Sapir and Whorf were interested in the power of multilingual awareness). Secondly, they shifted the inquiry from obligatory grammatical categories, such as tense, to lexical domains, such as color, that had a rather tenuous relationship to linguistic thought (color differentiation was, in fact, discussed by Boas and Whorf as an ability not influenced by language). Third, they shifted from concepts as interpretive categories to cognitive processes, such as perception or memory, that were of little interest to Sapir and Whorf, and proposed to investigate them with artificial stimuli, such as Munsell chips. In doing so, they moved the discussion further and further away from Sapir’s primary interest in ‘social reality’ and Whorf’s central concern with ‘habitual thought’.

When we look back, the attribution of the idea of linguistic determinism to multilingual scholars interested in second language learning and language change makes little sense. Yet the replacement of open-ended questions about linguistic diversity with two ‘testable’ hypotheses had a major advantage – it was easier to argue about and to digest. The transformation was further facilitated by four academic practices that allow us to manage the ever-increasing amount of literature in the ever-decreasing amount of time: (a) simplification of complex arguments (which often results in misinterpretation); (b) reduction of original texts to standard quotes; (c) reliance on other people’s exegeses; and (d) uncritical reproduction of received knowledge. Eventually, the very frequency of its reproduction made the SWH a ‘fact on the ground’.

Today, the received belief in the validity of the terms of engagement articulated by Brown and Lenneberg still reigns unopposed. Yet the focus on ‘non-linguistic cognition’ of ‘monolingual’ speakers in the experimental lab gave rise to a self-defeating line of inquiry that has little ecological validity and little in common with Whorf’s interest in thought insofar as it is linguistic. The purpose of The Bilingual Mind is to consider what is meant by linguistic thought and what non-trivial effects languages have on such thought in monolingual and multilingual speakers. Far be it from me, however, to claim that the book aims to move the inquiry on language and cognition ‘forward’. It does not – if only because I agree with Kuhn ([1962] 2012) that the metaphor of science as ongoing march forward is utterly misleading. My goal is to convince at least a few readers to move ‘away’ from the deeply familiar – yet inherently flawed – terms of engagement articulated for us by Brown and Lenneberg, to ‘backtrack’ towards the questions posited by Sapir and Whorf and to adopt more realistic terms of engagement with the relationship between language(s) and thought that take into consideration language change and the undeniable bi- and multilingualism of the majority of the world’s population.


The Bilingual Mind is due to publish February 2014.


Cambridge University Press announces launch of the Journal of Linguistic Geography

Posted on behalf of Editors William Labov and Dennis Preston


Cambridge University Press is pleased to announce the launch of the new online-only Journal of Linguistic Geography (JLG). The journal’s goal is to open the flow of linguistic analysis using electronic formats (such as scalable maps and figures, searchable data sets, and embedded audio files) in a field that has long been blocked by technical factors. For all new subscribers, a comprehensive User Experience Guide provides an overview of the journal’s interactive capacities. Submissions to the journal are welcome and may be sent to [email protected]. Queries are welcome, too.

The journal is an official publication of the International Conference on Methods in Dialectology. Editors Bill Labov (University of Pennsylvania) and Dennis R. Preston (Oklahoma State University) are supported by Technical Editor Bartłomiej Plichta (University of Minnesota). The full editorial board can be viewed here.

The Journal of Linguistic Geography: From Concept to Creation

The stacks of our libraries are filled with magnificent atlases of linguistic geography. File cabinets throughout the world are filled with papers that have never appeared, faced with the problem of reducing maps to small black-and-white versions that convey only a small part of the information in the original.

There will be no limit on the size of maps submitted to the Journal of Linguistic Geography; they will be viewed in their entirety with the panning and zooming options that are second nature to users of the internet. Color is as fundamental as size in cartography, and in electronic publication, color is no more difficult or expensive than black-and-white.

Even more crucial to analytical reading is the relation between map and text, which in print may require a back-and-forth paging operation that challenges memory and even lead to accepting (or rejecting) the author’s statement without making a point-by-point inspection. In the Journal of Linguistic Geography, maps and figures open in a new window, allowing the reader to make a direct comparison between what is said and what is shown.

A further advantage of the journal’s format is that of sound samples in the electronic page. They will not replace IPA notation, but rather serve to refine and encourage the use of phonetic notation.

Reading the Journal of Linguistic Geography will also show that technical innovations are not confined to modes of display. New developments in mathematical analysis of spatial patterns are represented and may include substantial appendices, since the space limitations of print journals do not apply.

So much for form. But what about content?

To put it simply, linguistic geography is concerned with the spatial differentiation of linguistic forms. Teachers of introductory linguistics find that students are fascinated with the fact there are regions nearby where speakers use ‘X’ to refer to what is (“rightly”) called ‘Y.’ This fascination with the facts of the matter impedes rather than encourages the development of our field as a branch of linguistic science. JLG hopes to mobilize those facts in pursuit of a better understanding of the nature of language structure and language change. Our interest is focused on those connections within language that reflect the impact of a given change on other members of the system. A submission that traces distribution of isolated forms or sounds will receive our full attention when it is woven into the fabric of relations that turn words into language.

We do not disprefer studies of the lexicon, but we encourage authors to display the use of a form against the background of competing and complimentary forms, showing what meanings are found for a given form as well as what forms are found for a given meaning.

Fields of structural relationships are most clearly delineated in phonology, and we would be surprised not to receive submissions dealing with the geography of chain shifts, splits and mergers, but we hope to deal with the geography of the full range of linguistic structures.

We invite studies of the perception of speech as well as production. We are interested in both how linguistic varieties across and within regions are heard and processed and how non-linguists perceive the spatial distribution of varieties, particularly when such studies shed light on the characteristics of language variation and change.

The fact that we are named the Journal of Linguistic Geography is not without significance, but the linguistics we appeal to is not just that of the internal relations of linguistic forms. It is also outwardly defined to include the social, historical and economic contexts in which language is formed and used. Thus we expect to find maps reflecting population growth and movement, out- and in-migration, political trends and voting records as well as highway and railroad networks.

Our Editorial Board comprises a group of distinguished linguists from throughout the world.  Learn more about these board members and how their own published work illustrates research of the scope and quality we hope to feature in the journal.

The Signs of a Savant

The Signs of a SavantWritten by Neil Smith, Ianthi Tsimpli, Gary Morgan & Bencie Woll

Every once in a while Nature gives us insight into the human condition by providing us with a unique case whose special properties illumine the species as a whole.  Christopher is such an example.  On first inspection his fate may not seem fortunate. Because he is unable to look after himself, he lives in sheltered accommodation; on a variety of standard tests of intelligence he scores poorly, with particular difficulty on non-verbal tests; his horizons seem to be limited to the performing of routine tasks of a non-demanding nature.  His life looks sadly circumscribed.  Until one turns to language.

Despite his disabilities, which mean that everyday tasks are burdensome chores, Christopher is a linguistic wonder: with varying degrees of fluency, he can read, write, speak, understand and translate more than twenty languages.  Playing noughts and crosses is beyond him, but interpreting between German and Spanish is easy; he doesn’t understand the kind of make-believe play that 3 or 4 year old children indulge in – pretending that a banana is a telephone for instance, but he learns new languages, from Berber to Welsh with enviable ease. His drawing ability indicates a severely low IQ of between 40 and 60 (a level hinting at ineducability), yet his English language ability indicates a superior IQ in excess of 120 (a level more than sufficient to enter University).  Christopher is a savant, someone with an island of startling talent in a sea of inability.

When you meet Christopher for the first time you are not sure of the best way to interact with him, as he is very shy but nonetheless interested in you. For anybody who studies language though, the way to communicate becomes immediately clear. You have to talk about language with him, not just what languages you speak but why you speak them, how you learnt them, how much you know them and what words you can share from those languages with him. For someone who loves language and the things you can do with languages it’s easy to talk to Christopher.

In an earlier book The Mind of a Savant we documented Christopher’s linguistic abilities in his first language, English, and his many ‘second’ languages, showing him to be an exceptional individual manifesting unusual and unattested asymmetries between abilities and deficits within and outside language.  In our new book The Signs of a Savant we revisit Christopher, elaborating on his obsession and talent for language, but concentrating on  how we taught him British Sign Language. BSL is a fascinating challenge for Christopher since it confronts his genius for language with a new modality which requires abilities where he is weakest.  He is mildly autistic, severely apraxic and visuo-spatially impaired, a combination which augurs poorly for his learning a signed language, which necessitates making eye-contact and the production and perception of fine motor differences of hand-shape and facial expression.

Somewhat surprisingly he wanted to learn to sign immediately even though he was faced with obvious barriers to picking it up: there are no books; you have to learn by looking into other people’s faces and you have to move your hands in complicated and coordinated ways. All of these factors caused Christopher problems but because the desire to learn about sign language burned so bright for him he slowly overcame these barriers and learned to sign. This book is about that journey and what we all learned from Christopher’s learning process: about language, about signing and ultimately about how the human mind works.

Although his production of BSL was considerably poorer than that of his spoken languages, his comprehension fell within the range defined by the students in the comparator group. Interestingly, he showed an asymmetry in his control of linguistic phenomena from the formal domain of language, such as negation, questions and agreement on the one hand, and his failure to master the classifier system – a domain where the control of topographic space is also required on the other. So our study sheds light on the similarities and differences between BSL and spoken languages: the similarities have to do with what is known as Universal Grammar – the language faculty as an autonomous component of human cognition; the differences have to do with the visuo-spatial Modality whose use represents a serious challenge to autistic individuals. At the same time, the asymmetries in his abilities provide evidence for the Modularity hypothesis of human cognition; the results of some of the tests we carried out support a particular theory of Memory, and Christopher’s case in general gives us insight into the nature of the human Mind.  We emphasize ‘human’:  despite the uniqueness of his case, Christopher is not a ‘Martian’. As we document in detail, the dissociations and asymmetries he manifests can be seen in other populations both pathological and typically developing.

If you want insight into a unique mind read The Signs of a Savant.

Out now in Paperback | 978-0-521-61769-7 | 232 pages | £21.99