Halliday and Language Development

May 4, 2010

Representing M. A. K. Halliday’s The Language of Early Childhood

In an attempt to understand how humans develop language, Michael Halliday spent twenty-one months studying early childhood language development. Although his sample size consisted of one child (his son, Nigel), Halliday emerged with insights that continue to generate meaning, nearly forty years later, across multiple fields and disciplines. Halliday worked as a participant observer, using pencil and paper to document Nigel’s utterances at six-week intervals from age 9 months to 2-½ years-old, with two points of focus in mind: the instance and the system. Halliday aimed to document Nigel’s progression as an individual, across each stage of his linguistic development; at the same time, he proposed a systemic theory of language that unites our earliest meaningful utterances with those we enact as adults. Through his three part stage-model, Halliday argued that humans develop language because we are creatures who need to mean, and language, above all else, is our primary resource for meaning.

For Halliday, language is always a resource for making meaning, and even the infant who cannot talk is developing language, and thereby, learning how to mean. Just as the infant can’t walk, but is learning how to use his body, he cannot talk either – at least not in the language of his mother tongue. Nonetheless, the child uses protolanguage (alternately referred to as protoconversation and protosemiosis) in order to express meaning, even before he has words in his communicative repertoire. His protolanguage, or child tongue, is created through interactions with native speakers of the mother tongue (i.e., caregivers, siblings, etc.). The child learns to mean through such occasions, and even though his protolanguage consists of basic content/expression pairs (not yet words, let alone higher order systems), his language nonetheless expresses meaning and performs concrete functions in the world. And while Halliday associates protolanguage with the child’s crawling stage, generally from six to twelve months, children voice meaningful intentions from much earlier in life.

For instance, Halliday cites an occasion when his twelve-day old son Nigel cried miserably until bath time, when his mother noticed “an unpleasant boil in the crook of his elbow” (vii, from Halliday’s personal notes). As soon as she called to her husband (Halliday), Nigel stopped crying, and did not cry again, even though the wound remained. Halliday considers this Nigel’s first act of communication: “he knew his mother had found out what was wrong, and that was what mattered” (vii). Thus, as early as weeks old, children (at least Halliday’s child, through his data) are communicating their needs and performing their desires. But so are our house pets, yet they do not develop language systems like we do. Indeed, Halliday considers Phase I of language development (i.e., the protolanguage stage) to be evolutionarily parallel to the language of pets. Nonetheless, this Phase I stage is the logical and necessary precursor to the lexicogrammatical Phase II, which bridges protolanguage and the mother tongue development of Phase III. For now, though, we will maintain our focus on Phase I, in order to consider the foundational role protolanguage plays in early literacy, particularly in terms of function and meaning-making.

For Halliday, Nigel’s early acts of communication are certainly meaningful, and yet they are not yet systematic. That is, they are not intentional choices of meaning expression, even if they are directed toward an addressee. As the brain and body develops, so does the child’s meaning potential through language. The child learns to move his head, to roll over, and correspondingly he becomes increasingly curious. When his caregiver interprets his squeaks and sounds semiotically, his communicative acts are given real-world value. In time, this confluence of experiences (i.e., brain and body development; expression/reaction occurrences) creates the conditions for stable, recognizable, even predictable signs, or content/expression pairs, to emerge. Protolanguage, like the physical act of crawling around, involves more systematically intentioned utterances, based on stable signs developing into systems. At this point, “meanings have become systemic. That is to say, each individual act of meaning is the instantiation of some meaning potential” (p. 12). This is key because meaning potential connotes real-world function, and if the function of language development is learning how to mean, as Halliday maintains, then language develops primarily to get things done in the world. This concept allows Halliday to holistically bridge a human’s language development from birth to adulthood, via the construction of language systems, each designed to facilitate meaning.

Even at the protolanguage stage (Phase I), language performs five primary functions (or microfunctions) across the following pragmatic contexts: instrumental, regulatory, interactional, personal and imaginative. That is, the child meaningfully voices his pre-verbal language to express particular intentions and desires. In order to “understand that which otherwise seems hard to pin down” Halliday managed to catalogue the various functions of these intentions, in order to “understand that which otherwise seems hard to pin down,” by asking, “what has the child learnt to do by means of language” (33, emphasis added). More specifically, this is what he found: the child enacts an instrumental function of protolanguage to obtain goods and services (these are “I want…” utterances); the regulatory (“Do as I tell you”) function aims to manipulate others; the interactional (“me and you”) function is the exchanging of attention; the personal (“Here I come/am…”) function enables the child to communicate his affective states; the imaginative function enables the child to pretend[1].

Again, the child is not yet using words, but is expressing intentions, thanks to his increasingly systematizing sign systems that correlate with his brain and body development. The child is using (proto)language to construct meaning as he constructs language from meaningful experiences, particularly by learning to distinguish when language and how language pays, in reality. Halliday believes that children are driven to expand their meaning potential in order to get more done through language. Thus, the need for words – in order to mean more effectively – is the driving force pushing the child into the lexicogrammatical stage, or Phase II, where words are first formed.

Before exploring the child’s development of words, it is time to give Halliday’s own words the floor, for this passage particularly illuminates his systems-based theory of language:

we can (now) gain a coherent picture of the child’s early semiotic development: how children are steadily increasing the number of ‘semogenic vectors,’ the various parameters that open up the total potential for meaning. First, they tease apart the content from expression; then they separate the system from the instance; then they open up further strata, further levels of organization within the content and within the expression; then they prise apart the distinct functional components inside each new stratum. With each step, they are opening up a new domain in which to move, so construing a multidimensional semiotic space analogous to the increasing dimensionality of the bodily space in which their material existence is located (p. 14).

Here again, Halliday correlates language, or rather meaning development, with body development, and the child’s potential for meaning expands as his environmental interactions expand. The brain, body, and environment, then, shape the child’s development (or meaning potential, or semiotic behavior). Additionally, we have considered how repeated instances become signs, which in turn become sign systems, and how selective and purposeful enactments of such now-systemic instances constitutes meaning-making language use (even in Phase I). Halliday asks us to consider the development of systems as a network, “and while each network taken by itself is a representation of just one ‘moment’ in the developmental progression, the sequence of several such networks presents a moving picture of the expanding consciousness of the child” (14-15).

When the body is ready[2], the child moves into Phase II, the transitional stage. This stage is distinguished from its predecessor because the child now speaks words and uses grammatical structures, but not in the sophisticated “plaurifunctional” way fluent speakers of the mother tongue do in Phase III. Instead, most new words perform only one function at a time. For instance, Nigel’s use of cat only meant “hello, cat” – an interactional function, which his use of syrup meant “I want my syrup” – an instrumental function (p. 40). In time, during Phase II, words begin to take on multifunctional properties, first over time, at separate instances, and eventually in a single instance. Halliday observed this when Nigel uttered the word cake and simultaneously meant “Look there’s a cake – and I want some!” (p. 41). Such a linguistic occasion, for Halliday, signifies an important advance toward adult-like speech, where words so often signify multiple meanings. Because function precedes vocabulary in Halliday’s model, instances such as Nigel’s cake utterance suggest two things: 1) vocabulary allows the child to expand his functional meaning potential; 2) vocabulary allows for the combination of functions.

Halliday refused to separate the learning of words from learning in general, and even though most of Nigel’s words during his transitional phrase did not serve multifunctional purposes (indeed they did not direct social interaction at all), they served the function of enabling learning. Such language use, for Halliday, is mathetic, as opposed to the pragmatic utterances of doing. In keeping with his systematized model of language development, he suggests that the mathetic function arises from a synthesis of the personal (self-oriented) and heuristic (other-oriented) non-pragmatic functions, and lays the foundation for communal and reflective behavior. Likewise, the child’s pragmatic utterances correspond to the action-based instrumental and regulatory functions of Phase I.

The child’s structures of language are “functionally specific,” early in the Phase II transition stage between child and adult language: they are either pragmatic or mathetic (p. 45) As the child moves closer to adult language, this functional binary shifts to a more abstract plane, where all adult language is inherently “plurifunctional” (p. 45). To explain this transition, I wish to present a claim of Halliday’s and then work backwards to unpack it. He says that, in grammatical terms, memory and self-consciousness:

constitute, respectively, proto-transivity and proto-mood; taken together, they make it possible for the child to transform experience into meaning – to reflect on and to act on the world and the people in it in one semiotic swoop (p. 20, emphasis added).

The ability to transform experience into meaning and to reflect and act on the world is characteristic of Phase III language development. That said, just as protolanguage precedes and enables the transition to adult language, Phase II precedes and enables the child to eventually “transform experience into meaning.”

But why are memory and self-consciousness so integral, and how do they develop? To answer this in Hallidayan terms is to begin to answer how the child is able to move into the mother tongue. This requires a return to the expression/content concept that enables signs, and then sign systems, and thus, protolanguage to develop. The twelve-day-old Nigel’s cries subsided once his parents located his wound. His cries, then, were an expression (signifier) of his physical discomfort, or the content of what his cries signified. This particular expression/content pair led to the parental response that satisfied the goals of the utterance. At twelve days old, a child is not making conscious expression/content pairings in order to procure distinct responses; however, the combination of responses (i.e., adult interpretations of child’s meaning) with expression/content pairings may eventually become schematic, to the point where the child understands that crying about a wound is a direct signal for attention and response.

I reiterate this point because Phase II signifies the structured formalization of content and expression in order to expand the child’s meaning-making potential. Content, or what is signified, is represented formally through lexicogrammar (i.e., the combination of words and grammar) and is related to semantics (or “meaning”). Expression, or the signifier, is represented formally through phonology and is related to phonetics, and thus, the body. Together, the content and expression levels of signification interface between the formal systems of language and the material world of the child’s environment. This is why every adult utterance is multifunctional – because such conditions are built into the grammar – and it is why adult, higher-order language enables infinite meaning-making possibilities.

On the expression plane, the body is physically ready to enable phonological (as well as non-verbal) development, and at the same time, the child begins making sense of his world on the content plane. Both of these also enable the child to enact his own “social being” with others, with real and immediate effects. As the child moves from “the primary semiotic of the protolanguage to the higher-order semiotic of the mother tongue,” memory and self-consciousness begin to develop. Memory allows the child “to construe classes of phenomena out of repeated instances, using a re-entrant mapping to impose categories on its experience of the world” (p. 20). This means that the child can construe common nouns, which is the beginning of referential meaning. Self-consciousness enables the child to distinguish between declarative (“how things are”) and imperative (“how I want things to be”) meanings. Together, in grammatical terms, memory and self-consciousness create the conditions necessary for a child “to transform experience into meaning – to reflect on and to act on the world and the people in it in one semiotic swoop (p. 20). But again, this cannot happen without the increasingly specialized bodily domain of expression. For Halliday, language development, or learning how to mean, through body, words, and grammar, is a reflection of brain development.

Another proto- element of Phase II is that of dialogue, which functions through the adoption and assignment of social roles. While protolanguage suggests dialogic possibilities (e.g., Nigel’s cries of pain; his parents’ response), the dialogue of adult language “exemplifies the general principle whereby people adopt roles, assign them, and accept or reject those that are assigned to them” (p. 46). Halliday notes that Nigel develops the ability to respond to Wh- questions in Phase II; that is, he is able to accept the roles he is assigned but is not yet able to assign roles to others. Nonetheless, his protodialogue suggests he is on his way to achieving the most sophisticated of early language functions: the informative function. Dialogue also enables the child to develop mathetic and pragmatic (especially personal, interactional, and heuristic) functions, which allow for manipulations of mood (declarative, interrogative, etc.). In turn, this opens up “the entire interpersonal component in the language system,” which is where the speaker “intrudes or…builds himself into the language structure, expressing his relations with other participants, his attitudes and judgments, his commitments, desires, and the like” (p. 48). Ultimately, this is how the child participates in and receives the culture in which he operates.

According to Halliday, “Phase II can be said to end when the child has mastered the principles of grammar and of dialogue, and thus effectively completed the transition to the adult language system” (p. 49). The adult language, on a functional level, is comprised to two metafunctions – the ideational and interpersonal – and the textual function, each of which maps outward from Phast I and Phase II components. [Halliday called it systemic functional linguistics for a reason.] Each of these functions represent major systems, with distinct linguistic properties and options, within the larger language constellation. The ideational system embodies the speaker’s (Halliday shifts from “child” to “speaker” in Phase III) personal experience and interpretation of his internal and external realities; the interpersonal system enables the speaker to assume and assign social roles, to both observe and intrude; the textual function enables the construction of texts[3], which can be inhabited by the other two metafunctions.

Where once function could be tracked through simple content/expression pairings, Phase III-level function is significantly more complex, and yet, in the scope of this theory, is nonetheless an explosion of the primary content/expression dynamic. Typical utterances in adult language have both ideational and interpersonal meaning-making components. Halliday invokes functional grammar to point out that every main clause is a structured instantiation of both mood (interpersonal) and transivity (ideational). Through his intense study of Nigel, motivated by the ideas of Bernstein, Firth, and the many frustrated teachers of language he worked with, Michael Halliday aimed to construct a unified, holistic theory of language development. By 1975, he succeeded at least to the point where he could write the following:

It appears, then, that the ‘metafunctions’ of the Phase III grammatical system arise, indirectly but unmistakably, out of the primary uses of language that the child develops in Phase I. . . . Hence the child’s Phase I functional system, which is a system of the content in a “content/expression” language, evolves along the familiar lines of generalization followed by abstraction into the Phase III (adult) functional system, which is a system of the form in a “content/form/expression” language. . . . the original Phase I functions have not just disappeared (but) have become the generalized contexts of language use (p. 55).

Imagine the satisfaction he must have felt when composing that passage! But Halliday was not even close to finished. Some of his most exciting work concerns those “generalized contexts of language use” (i.e., the adult language abstractions of primary human functions).

In “The Social Context of Language Development,” Halliday (1975b) examine the intersection of two meaning-making (semiotic) systems: culture and language, primarily at the level of text. Text, he says, is any language – speech or writing – that is operational, and therefore means. Text is not simply sounds, letters, and words; it is choice incarnate. Text is not only what is but is “set against the background of what might have been” (p. 285). The multitude of choices that comprise texts (Halliday calls these microlinguistic acts) represent the actualization of meaning-potential, on an individual, social, and cultural level. These microlinguistic acts can be traced back to infancy, when the child first learns the value of content/expression pairings, whose functions are activated by adult-level respondents in real-world contexts.

From the start, meaning is constructed through language that arises from situations (e.g., Nigel’s wound). As these meaning-making situations become typified, content/expression pairs become encoded as signs, and the language building process is underway. In yet another echo of protolanguage in the world of the mother tongue, Halliday identifies situation type as “a generalized context of a situation” and argues that all text is indeed situated (p. 287). This is because texts by definition are functional, constructed to serve social purposes, by humans intent to mean. The development of language, after all, is learning how to mean.

Texts then are situated, both linguistically and culturally, and while they consist of a multitude of microlinguistic acts (i.e., enacted options), they are shaped the conditions of particular situation types[4], including the typified conventions they enable, as well as the linguistic capacities and roles of the participants involved. For instance, a typical cultural situation type may be that of “mother reading bedtime story to child.” The situation, which is clearly a type (virtually everyone can conjure it), has served a social, cultural, linguistic, and functional purpose for generations. The situation type itself is woven into the fabric of our culture to the degree that it enacts participants as participants enact it. In this way, participants are not overwhelmed by the need to create language and communicative roles from scratch each time they interact. The roles themselves, as well as an array of microlinguistic acts, are built in to the situation; therefore, as language learners conceptualize situation types, they also gain access to the array of discourse conventions embedded in these situations.

Halliday stratifies situation types along textual, linguistic, and social planes, but these are best captured by his notion of register. If situation types are typified social contexts, register is characterized by the typical range of meaning-making options that are available within the situation. Halliday identifies three components of register – field, tenor, and mode – which can be used to analyze, understand, and represent the situated, meaning-making activity of any communicative act. The communicative field, which corresponds to the ideational metafunction (i.e., transivity), concerns what takes place, and incorporates setting and subject matter. Tenor corresponds to the interpersonal metafunction and concerns the participants, their tone, and the discourse styles afforded and influenced by their role-relationships. Mode represents the textual function and concerns the symbolic or rhetorical channel by which language travels, including written or spoken genres and, indeed, speech or writing itself. Collectively, these components of register create the conditions for communication, and thus, for the creation of meaning through language.

Roughly twenty-five years after identifying the seven primary functions of early childhood language, Halliday (1998) remarked that recording and interpreting (i.e., representing) the child is easy, requiring little more than pen and paper – as long as the child has not surpassed the protolanguage stage. At this point, his behavior is not yet systemic, so what you see is what you get. Representing the older child, even as young as three years of age, is significantly more complex. This child’s language is “fully systemic,” consisting of a system of networks, each stratified along cultural, social, grammatical, phonological, and situation planes. Halliday’s grand theory attempts to unite primal, infantile utterances with the sophisticated language exchanges of adulthood by arguing that language development is motivated above all by a very human need to make meaning. And it makes me wonder: to what extent is the composition of this paper merely an extension of my first days; being human, after all, is a situation type.


Halliday, M. A. K., (2003). The Language of Early Childhood. Volume 4 in the Collected Works of M. A. K. Halliday, ed. by J. Webster. London: Continuum.

Halliday, M. A. K., (1998). “Representing the Child as a Semiotic Being (One Who Means)” In The Language of Early Childhood.

Halliday, M. A. K., (1976). “Early Language Learning: A Sociolinguistic Approach” In The Language of Early Childhood.

Halliday, M. A. K., (1975). “Learning How to Mean” In The Language of Early Childhood.

Halliday, M. A. K., (1998). “The Social Context of Language Development” In The Language of Early Childhood.

[1] Halliday is less than clear about whether the imaginative function is part of Phase I or Phase II, but he does say this: “the first four functions listed clearly precede the rest,” which in addition to imaginative includes heuristic (“tell me why”) and informative (“I’ve got something to tell you”) functions.

[2] “The brain is no more capable of construing a grammar, at birth, than the human body is of walking – or the brain of directing it to do so” (p. 18)

[3] Modern genre theorists have argued that such a construction is a two-way street, where texts invent users as they are invented by users

[4] This is a foundational principle of rhetorical genre theory.


5 Responses to “Halliday and Language Development”

  1. Ant said

    He’s an interesting cat. And this was an honest effort.

  2. Zayn said

    This is really good.it will contribute imensely to the field of psycholinguistics

  3. Tamoka said

    Thank you for posting this, I have found it better explained than the original!

  4. Harrington said

    I wanted a list of phases as listed by Halliday no additions or subtraction.thank you

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: