Recommending Monolingualism to Multilinguals – Why, and Why Not

Multilingual christmas lights in Barcelona

Photo by

In cases of suspected or confirmed clinical disorder among bilingual/multilingual children, one common recommendation is to have the children “switch to one language.” This advice comes both from monolingual SLPs, who are trained in and for monolingual settings, and from multilingual SLPs, including those working in multilingual contexts. I would like to offer a few thoughts on the practical feasibility of this advice, the reasons that may motivate it, and whether those reasons match what we know about multilingualism and speech-language disorders.

Recommending monolingualism to multilinguals seems to draw on a conviction that multilingualism either causes or worsens speech-language and related disorders or, conversely, that monolingualism either blocks or alleviates them. Speech disorders (such as stuttering), language disorders (such as SLI), and developmental disorders (such as autism) do affect language, in that linguistic development relates to physical, cognitive, social and emotional development. But language development can be typical or atypical regardless of the number of languages in a child’s repertoire. Speech-language and developmental clinical conditions affect multilinguals and monolinguals alike, which means that there is no correlation between multilingualism, or monolingualism, and disorder. In the absence of a correlation, there can be no legitimate conclusion that using one language vs. using more than one has predictable effects upon disorder. The unwarranted conviction that number of languages is a relevant factor of speech-language disorder rests on a number of beliefs, as follows.

First, the belief that healthy linguistic and related development can only be achieved in a single language. Multilingual children naturally develop linguistically in all the languages that they need to use for everyday purposes. Cognitive, social and emotional development follows suit, through each of the contexts in which the languages of a multilingual are relevant. Multilinguals, big and small, use each of their languages in different ways. This is in fact why they are multilinguals: if a single language served all their purposes, they would be monolinguals.

Each of the languages of a multilingual naturally reflects the specific uses that it serves, and each will develop accordingly, at its own pace. If a child uses, say, one language with mum, another one with dad, and yet another one in school, each language will naturally show evidence of mum-related, dad-related and school-related accent, vocabulary, grammar and pragmatics. Having different words or a different number of words in each language, for example, or preferring to use one language rather than another for specific topics or with different people, is typical of multilingualism, not a sign of atypical linguistic competence. A less developed language of a multilingual is therefore not a symptom of a clinical condition such as ‘language delay’, but reflects instead less use of that language than of another. If there are concerns about the development of a particular language of a multilingual, the child may be appropriately referred to a language tutor, not to an SLP.

Second, the belief that using more than one language results in diminished proficiency both in each language and in other proficiency. This belief draws on subtractive views of the human brain, which have it as a computer-like processor featuring limited storage capacity, organised into computer-like modules and processing modes. On this view, ‘brain space’ allocated to each language disrupts other brain space, by encroaching upon it in ways similar to zero-sum situations, where the gains and losses of one ‘module’ exactly match the losses and gains of another, respectively. Computer analogies of the human brain gained popularity by the middle of last century, but current findings about inherent brain plasticity prove their inadequacy to model brain organisation, activity and power.

Third, the belief that using one particular language in one setting will promote development of that language in other settings. The recommendation to switch to one language often means ‘switch to exclusive use of the mainstream language at home.’ Even in cases where it might be viable to change or amend the home language practices in which a child has been brought up, switching to a mainstream language at home, or making it the only home language, will not necessarily impact uses of that language elsewhere, for example in school. The converse is also true: academic uses of a language, say, do not automatically transfer to home uses of the same language, because these uses belong to different registers.

“Register” is a term used in linguistics to describe the differential ways in which we all use our languages to fit specific contexts and specific people. Monolingual children (and adults) switch among the registers that they have learnt to be appropriate at home, in school, at work, or with peers, juniors and elders. Multilinguals do likewise: they switch register in each of their languages, in order to match the participants and the context of an interaction in a particular language, and they switch language, again where participants and context so require. The ability to switch uses of language appropriately constitutes proof of linguistic competence, because it shows understanding of how different registers and/or different languages serve different purposes. A home language, or a home register, develops for home-use purposes, which do not and cannot match academic and other uses of it. The way to promote development of languages or registers in a specific context is to use them in that context.

Finally, the belief that language disorder is best addressed through a single language of intervention. The mainstream language favoured by recommendations of monolingualism often coincides with the language of education, that the child may, in addition, happen to share with the clinician. This raises the question of whether the recommendations are indeed meant to favour monolingualism, or to favour monolingualism in a particular language, the language in which assessment instruments are likely to be more readily available. Whichever the case may be, current research on clinical work with multilingual children shows that intervention which targets the whole of a child’s linguistic repertoire increases both the chances and the pace of recovery. Addressing linguistic repertoires for purposes of intervention makes good overall sense, in that language disorders affect the whole of a child’s linguistic repertoire, regardless of the number of languages involved. Diagnosis must take the whole child into account, so that intervention can start from where the child’s abilities are, whether these abilities are monolingual or multilingual.

Depending on the context of specific interactions, typical monolinguals and multilinguals alike make proficient use of their linguistic repertoires, which means differential use of linguistic resources. The whole linguistic repertoire of a monolingual child translates into resources drawn from a single language, but the whole linguistic repertoire of a multilingual child does not. Beliefs and convictions to the contrary, such as the ones sketched above, rest on a misconception of monolingualism as “norm” of language use, which has spawned related misconceptions that take proficiency in a single language for linguistic health, and lack of proficiency in a single language for symptom of language disorder. Being multilingual involves differential proficiency in more than one language, whose interplay with social, cognitive and emotional development can only be ascertained from observation of the child’s abilities in each appropriate context.

The take-home message that I would like to leave here is that multilingualism is neither a disorder nor a factor of disorder. In cases of suspected or confirmed clinical disorder among bilingual/multilingual children, switching to a single language will not address the disorder. It will simply create a monolingual child with a disorder.


Madalena Cruz-Ferreira, PhD in Linguistics and Phonetics (University of Manchester, UK), researches multilingualism and child language. One section of her book Multilingual Norms addresses multilingual clinical assessment. Her blog Being Multilingual deals with the use of several languages at home, in school and in clinic.

Multilingual Typicality vs. Speech-language Disorder

table with coffee mugs and maps

Photo by minka6

Any assessment involves a comparison. For assessment purposes, we use comparison always one way. We compare X to Y, never Y to X, because we have satisfied ourselves of two conditions: first, that Y is a reliable benchmark, which specifies a particular norm of behaviour, including linguistic behaviour; and second, that the behaviour of X can be fairly assessed through the use of that benchmark.

Reliable benchmarks are norm-referenced and standardised for particular populations. Since different populations use different languages and different varieties of the same language, we seek to provide ourselves with developmental and/or clinical assessment instruments which are normed accordingly. We know that it would be as unfair to test, say, users of Korean with instruments normed for Portuguese as to test users of Canadian French with instruments normed for Belgian French. Although we still lack normed instruments for most languages and language varieties, limitations imposed by assessment in these less than ideal conditions are well understood. In monolingual settings, with monolingual clients, clinical practices take them into account.

Concerning multilingual clients, however, the situation is quite distinct. In what follows, I take the words multilingual and multilingualism to refer to users/uses of more than one language, that is, to include bilingual(ism), trilingual(ism), and so on.

One first observation is that multilinguals stand for a disproportionate number of referrals to both special education and speech-language therapy, compared to monolinguals. We may start by asking ourselves why multilinguals are consistently compared to monolinguals, but not the other way around. The reason is that monolingualism has been assumed as a norm of linguistic usage, which has besides become synonymous with cognitive, social and linguistic health. The reason for this, in turn, is that the first researchers who addressed multilingualism were monolingual, or subscribed to monolingual approaches to language, or both. The tradition of thought that they initiated almost one century ago lingers on, and shapes the many misconceptions surrounding multilingualism. A few examples follow, showing how these misconceptions are interrelated and entail one another:

  • “Multilinguals are special.” In monolingual countries and settings, multilingualism is viewed as the special case of language uses. Clients who are multilingual are labelled as such, whereas clients who are monolingual are not labelled as monolinguals. Given that multilinguals outnumber monolinguals worldwide, it cannot be the case that the majority of the world’s population is “special”. The century-old tradition that takes multilingualism as special started by also taking it as the correlate, and sometimes even the cause, of diverse cognitive, social and linguistic shortcomings. The current emerging trend, that lauds multilingualism as unquestionably positive, simply perpetuates the (mis)perception that multilinguals are “special”.
  • “Multilingualism means equivalent proficiency in all languages.” This assumption is better described by a term that I coined, multi-monolingualism, to label the underlying belief that a multilingual equals several monolinguals. This is not what multilingualism is. If multilinguals could (or should) use all their languages in exactly the same way, they would not need several languages: one all-purpose language would be enough. “One all-purpose language” defines a monolingual, not a multilingual. Multilinguals use their languages in different ways, with different people, in different situations, for different purposes. This is why their languages develop differently and cannot therefore be made equivalent.
  • “Ability in one language reflects language ability.” Clinical findings about one of the languages of a multilingual client are often taken as a reliable reflection of the client’s overall language ability. “Language ability” concerns the whole of an individual’s linguistic repertoire, not ability in a particular language. Taking the one for the other means taking a multilingual for a monolingual. The full linguistic repertoire of a monolingual does consist of a single language, but the full linguistic repertoire of a multilingual does not.
  • “Multilinguals can be fairly assessed through monolingual instruments .” The assessment instruments that are  available to us so far are monolingual, and naturally reflect monolingual norms. In addition, multilinguals tend to be assessed either in mainstream languages, or in languages for which assessment instruments have been standardised, neither of which may accurately portray the clients’ linguistic ability. In the absence of normative guidance about multilingualism, skewed findings about multilingual behaviour are to be expected. False positives, where typical multilingual behaviour is mistaken for disorder, account for the disproportion of referrals mentioned above. But, equally seriously, false negatives mistake disorder for typical multilingual behaviour, and so fail to identify disordered multilingualism.

Ideally, then, we should provide ourselves with standardised instruments devised for multilingual uses of language, based on multilingual norms of usage. These norms are not, as I hope to have made clear above, “multi-monolingual”: there are typical behaviours among multilinguals, just like there are typical behaviours among monolinguals. The difference is that we have failed to pay attention to the former, because we have taken the latter as the benchmark of linguistic behaviour across the board. The issue here, as always, is that without knowing what is typical, we cannot tell what is deviant.

Current developments, which take a fresh look at multilingualism, from a multilingual perspective, already show promising results. One example concerns mixes, the use of features of several languages in the same utterance or exchange. Mixes have been stigmatised as instances of “semilingualism”, whereas they are a multilingual norm of usage. The regularity of mixed patterns in typical multilingual speech has been found to aid in the diagnosis of SLI (specific language impairment), in multilingual children whose mixing patterns deviate from the norm. Another development concerns the use of what is known as dynamic assessment, in clinic. Dynamic assessment methods involve teaching and testing linguistic items and structures that are independent of particular languages, and that therefore probe for language ability, not ability in particular languages.

Growing awareness about the lack of multilingual norms also impacts the clinicians themselves. To the best of my knowledge, professional training of SLPs does not include information about languages other than the language of intervention, or about multilingualism itself. This is so even for multilingual SLPs, or for those who plan to practise in multilingual settings. Many SLPs thus encounter multilingualism for the first time in clinic, where the “special” status accorded to multilinguals may well shape expectations about multilingual clients. There is of course no requirement that SLPs become multilingual. Being multilingual does not mean understanding what multilingualism is: misconceptions about multilingualism are shared by monolinguals and multilinguals alike. The requirement is that SLPs, and the rest of us, become familiar with what multilingualism is, so we satisfy ourselves that, while we wait for the standardisation of multilingual norms, we are giving multilinguals a fair assessment chance.

Multilingualism is not about what several languages can do to people, it is about what people can do with several languages. The same can be said about monolingualism and a single language: the number of languages that people happen to need to use in order to function appropriately in their everyday environments has little to do with their language ability, just like the number of musical instruments that one plays has little to do with one’s musical ability.

One final note: I have discussed multilingual assessment in my blog, which is geared to a general audience, in a post titled The fight for a fair deal. For more specialised research and findings on multilingual typicality, my book Multilingual Norms may be of relevance.

Madalena Cruz-Ferreira, PhD in Linguistics and Phonetics (University of Manchester, UK), researches multilingualism and child language. One section of her book Multilingual Norms addresses multilingual clinical assessment. Her blog Being Multilingual deals with the use of several languages at home, in school and in clinic.