Deep Language, Intelligence and Language Proficiency in Learning

Proceedings of the Linguistic Society of South Africa conference. (University of Port Elizabeth, South Africa, 1995).

Author – Raphael Gamaroff

See also: Language as a deep semiotic system and fluid intelligence in language proficiency, and Cognition & Language proficiency













The arguments in this article are based on the premise that intelligence plays a major role in learning. It is argued that intelligence is a “deep” language, which, although similar to language proficiency in some respects, is fundamentally different in others. Oller distinguishes between “deep language” and natural languages (e.g. English, Zulu, Russian). Oller argues that the reasoning processes involved in “deep language” are the primordial mental processes involved in all learning. Cattell’s Culture-Fair Intelligence Test (CFIT) is examined as an example of a test of these “deep language” processes.

Oller also uses the term “language as intelligence”, which he regards as equivalent in meaning to “deep language”. With regard to language (proficiency), Oller (following Cummins) distinguishes between Basic Interpersonal and Communicative Skills (BICS) and Cognitive and Academic Language Proficiency (CALP). It is argued that Oller’s “language as intelligence” is a confusing concept, because it doesn’t distinguish between the “general intelligence” common to all human being and the “intelligence” that is required for academic success. It is only by taking this distinction into account that one can appreciate the further distinction between the automatic acquisition of BICS and the conscious learning of the more intellectually sophisticated CALP, where the latter plays a crucial role in academic learning. A distinction is made between “thought” and “intelligence”, which helps to clarify the notion of “language as intelligence”. An understanding of the distinction between BICS and CALP and their relationship to “intelligence” is indispensable in understanding academic learning.


Although there is a great need in applied linguistics to “bridge the gap between linguistic theorists and language practitioners” (Young 1990:6) by relating “our practice to the base discipline of linguistics” (Young 1990:10), there is something else that should be seriously considered by applied linguists, namely, the relationship between language and cognition. This relationship is the topic of this paper. Young (1993:276) makes the following important point in this regard:

Applied linguistics, if it is to be relevant in this process [i.e. the reconstruction of education in South Africa], needs to depend less on social theory for its inspiration and descriptive and explanatory power. We need to explore the deep structures of language and cognition more seriously than we do now. [My square brackets]

I agree with Young and would therefore like to explore the “deep structures of language and cognition”. In so doing, I shall try not to get trapped in common sense judgements. According to Chomsky (1972:26):

We must recognize that even the most familiar phenomena require explanation and that we have no privileged access to the underlying mechanisms, no more so than in physiology and physics. Only the most preliminary and tentative hypotheses can be offered concerning the nature of language, its use, and its acquisition. As native speakers, we have a vast amount of data available to us. For just this reason it is easy to fall into the trap of believing that there is nothing to be explained, that whatever organizing principles and underlying mechanisms may exist must be `given’ as the data is given. [For Chomsky “data” is a singular noun]

This “vast amount of data available to us” (Chomsky above) consists of natural languages such as English, Tswana, and Russian as well as artificial languages such as formal logic, mathematics and computer languages. But, there may also exist another kind of language that could be the source of both natural and artificial languages. This possible kind of language has been given many names, e.g. “deep language”, “language of thought” (referred to henceforth as LOT), “language as intelligence”, “mentalese”, a “deep propositional reasoning system”. I discuss the relationship between “deep language” and natural language, and thus exclude the relationship between “deep language” and artificial languages. Suffice it to say that artificial languages cannot develop without language: natural langauge and deep language.

Cattell’s Culture-Fair Intelligence Test (CFIT) is used to show the connections between non-verbal intelligence and language proficiency. I examine the possibility that this non-verbal intelligence is also a language; a deep language, which has strong connections with language proficiency and learning. Such an analysis may help bridge the disturbing gap, which Young (1990) complains about, between applied linguistics, pure linguistics and psychology.


The complexity of the relationship between language and thought continues to frustrate (violate?) the ablest minds. Indeed, for Groddeck (1977:249-250), language is a violation of thought; a lie:

We are thus faced with the fact that every word in our language is a lie, whether it is uttered by our mouth or remains mute in the brain, a lie which violates facts, which makes us look at the world from a false perspective and think falsely.

In contrast to Groddeck’s negative view, Vygotsky (1962) sees thought and language in terms of a symbiotic relationship, where “a word devoid of thought is a dead thing, and a thought unembodied in words remains a shadow” (Vygotsky 1962:153). For Vygotsky, language is thought’s crowning glory; for Groddeck, its crown of thorns.

Devitt and Derelny (1987:127) maintain that language originated out of a need to understand the environment and ourselves in order to use and control the environment. Out of this need to control the environment culture was born. Primitive man conveyed meaning through body language such as grunts and gestures. Grunts and gestures caught on out of which was born linguistic conventions. The capacity to think – according to Devitt and Derelny – is borrowed from those who created these conventions and thus primitive thought was made easy. (It is not explained who these “creators” of conventions were). The drive to understand leads to more complicated thoughts, to more complicated speaker meanings to more complicated conventions. “If this sketch is right, they add, we have, as individuals as a species, engaged in a prodigious feat of lifting ourselves up by our own sematic bootstraps”. Thus, through the introduction of a public language, the “language of thought” (Devitt and Derelny, 1987:127) is able to expand into a discourse. This description of language development fits in very well with Vygotsky’s view (and I would think with other authors such as Hayakawa and Hayakawa 1990; and Halliday 1977).

With regard to language, Vygotsky (1978:28) maintains that human beings would be no different from animals without the social contact with other people. This is indeed so. But not because the social imprints itself onto a tabula rasa, but because the human brain has the innate structures to learn, whether it be a language or some other kind of “tool”.

Vygotsky adopted Engels’ notion that labour and tool use transforms human nature. Vygotsky “celebrates the social formation of the mind” (Daniels et al. 1995) through the mediation of the supreme mental tool, namely language. According to Gregory, we are “forced to attribute the way we think very largely to our tools and what has been created by them” (Gregory 1981:42). This means that tools (technology) would be the source of the most abstract notions of philosophy and theories of mind, and science.

Although there is no doubt that a psychology that is merely based on biology is inadequate, this does not mean that tools are the only source of the abstract notions of the brainmind. Surely there is a symbiotic relationship between the development of the brainmind and mind tools such as tools of measuring, calculating and thinking and communicating. The tool of tools (and the tool of fools) is the sign:

The use of signs leads humans to a specific structure of behaviour that breaks away from biological development and creates new forms of a culturally-based psychological process (Vygotsky 1978:40).

It is Vygotsky’s term “breaks away” (which I assume is a correct translation of the Russian text) that I find problematic. In order to explain the problem, it would be useful to examine the concept of maturation. According to Vygotsky, it is a fact that “maturation per se is a secondary factor in the development of the most complex, unique forms of behaviour.(Vygotsky 1978:19). Secondary? I don’t think so. To explain why I disagree with Vygotsky, I shall look at language in the context of growth and maturation. In the discussion that follows I shall not refer to Piaget’s views (which are well known by most social psychologists) but to Eric Lenneberg’s views on the biological basis of language (which are very similar, if not identical, to Chomsky’s views on the same topic). I present the salient features of Lenneberg’s biological view of language development:

Language for many seems to consist of arbitrary cultural conventions, e.g. Wittgenstein’s “language games”. However, fruitful explanatory principles are of a biological nature.There are radical differences between the rules of games (arbitrarily determined) and the rules of language (biologically determined). However, we can still speak of the biology of game-playing (what are the capacities that enable a human to play a game of chance, to “waste” one’s time away. An animal (humans are included under “animal) is not like a tool that can be arbitrarily assigned a particular use. The structure of the brain and the body are an interdependent unit. These structures are programmed from within, i.e. genetically (Lenneberg 1967:1-4). Thus the developmental process is physiological in nature. Modifications after birth are determined by genetic and prenatal events. The degree of plasticity (e.g. the diversity between individuals) is the product of biological conditions. Specificity and plasticity are mutually independent and all depend on the environment. “Thus the notion `dependence on environment’ (which by implication is the same as `dependent on experience’) is not a useful criterion for the classification of behaviour (Lenneberg 1967:10-12). In other words, culture is bioculture (where culture is subsumed under the notion “environment” in this article); and sociology is biosociology (Boyd and Richerson, 1985). There is not only a logic of biology, but also a biology of logic – this circular description illustrates the basic epistemological and hence ontological problem of the mind-brain duality, which is ultimately a problem of the origin of logic – the logos, the “word”.

Advances in brain sciences and genetics provide additional support to the theory that “deep language-like representational systems are critical to all aspects of biological organisation and neurological functioning” (Oller, 1991:4). There may indeed be more in the biblical proposition “In the beginning was the word” than meets the ear. For Vygotsky (1962), on the other hand, “in the beginning was the deed”, and “the word is the end of development, crowning the deed” (quoted from Oller, 1991:35; see Note 1). It was only a few months after Vygotsky had made the last remark that he died “never dreaming that his remarks would be tested in scarcely three decades by one of the most remarkable advances in the history of science – the discovery of the genetic code” (Oller, 1991:35), and also by the remarkable advances in the brain sciences (Clark, 1977; Danesi, 1994; Paradis, 1991; Perecman, 1989; Pribham, 1971; Woese, 1967:4; Young, 1978). Vygotsky got it wrong; the word, the “deep’ word, is the beginning not the end of development. The constraints of this paper do not permit a detailed discussion on the biological evidence for “deep language”. Oller (1981) and especially Oller (1991) provide a good summary of theories on the biological foundations of deep language (that is, LOT).


Some authors argue that it is possible to think without (natural) language – indeed that if one couldn’t think without language, one would not be able to think – or use language – at all.

For example, Bloch (1991) maintains that much of our thinking is done without language. Bloch’s position is supported by Fodor, Bever and Garrett (1976), Fodor (1986) and Oller (1981), who maintain that there is a non-verbal “language of thought” or “deep language”.

Oller (1983a:355; see also 1981) defines deep language as a “deep propositional reasoning system”. Fodor, Bever and Garrett (1974:375) describe the role of this “deep propositional reasoning system” in the following way (this would be Oller’s view as well):

Deciding on an action is, among other things, a computational process. In particular, it presupposes that the agent has access to a system of representation in which the various behaviourial options can be formulated and assessed…Deciding upon an action itself involves the use of a language-like system, and this is true whether or not the action up for consideration happens to be a speech act.

It is this “deep propositional reasoning system” that is claimed by the above authors to be the driving force behind all thought, whether the thought involves high level reasoning tasks such as doing IQ tests (which will be discussed shortly) or low level reasoning tasks such as tying shoelaces.

Let us now consider whether there is a difference in meaning between the “deep propositional reasoning system” described by Fodor, Bever and Garrett and what I consider to be a conventional definition of reasoning. Consider Carroll’s explanation of reasoning (1964:93):

Thinking aided by [natural] language is called reasoning, and the ability to reason depends largely on the ability to formulate steps in an inferential process in terms of language. [My square brackets]

For Carroll, reasoning only comes into play when “thinking [is] aided by language”; but for Oller (1983a:355), the mind’s “deep propositional reasoning system” precedes “language” (natural language). In other words, for Oller (and for Fodor, Bever and Garrett), one can reason – in the “deep” sense of the term – without (natural) “language”. For Carroll, thinking is inchoate, until it is “informed” by language, whereas for Oller, the “deep propositional reasoning system” itself is a component – indeed the principal component – of thinking. (My underlining and square brackets). Thus, for Oller, there are two levels of reasoning; the deep level and the level that Carroll describes.

I need to point out that this “deep propositional reasoning system” is not identical to the “deep structure” of Chomsky’s transformational grammar, which pertains to the grammar of a particular language (i.e. a natural language like English or Tswana) and designates a particular stage in the derivation of a sentence. I cannot be sure from Chomsky’s writings whether his “universal grammar” only pertains to a metatheory or schematism for grammar (natural language) – LAD – or whether it is equivalent to LOT. Fodor (1976), on the other hand, seems to equate “universal grammar” with LOT. Young’s meaning of “deep structure” in the introduction to this paper is used loosely and does not refer specifically to LAD or LOT. (Strictly speaking LAD and LOT should be preceded by “the” – but I prefer to leave it out). (See criticisms of the notion of LOT [Loevinger and Rey 1992; Nolan 1993] and rebuttals of some of these criticisms (e.g. Fodor in: Loevinger and Rey 1992).


In this section, I examine Cattell’s CFIT, which Oller uses to illustrate what he means by a “deep propositional reasoning system”, i.e. deep language, mentalese, or the language of thought.

Cattell’s CFIT (1973:7); see also Oller, 1981:484) consists of three scales (i.e. levels):

Scale 1: Four to eight years of age.

Scale 2: Eight to 14 years and adults of average intelligence.

Scale 3: Adults of high intelligence.

Each scale consists of four subtests, which are composed of visual symbols that involve the perception of relationships: 1. Progressive series completion, 2. Classification, 3. Matrices and 4. Conditions. Sample items of Scale 2 are provided in Figure 1:

Figure 1

Cattell’s CFIT

Sample items of Cattell’s CFIT

1. Progressive series completion

2. Classification

3. Matrices

4. Conditions

These subtests test the following four operations:

Subtest 1. Progressive series completion. In the example above, the bar becomes progressively longer. The correct answer is choice 1.

Subtest 2. Classification. Five figures are presented where the aim is to select the one different from the others. The correct answer is choice 4.

Subtest 3. Matrices. The aim is to complete the matrix presented in the bottom right hand corner. The correct answer is choice 1.

Subtest 4. Conditions. The aim here is to select from the five items the one where the dot would lie outside the box but inside the circle. Only choice 3 meets these conditions.

With regard to Cattell’s subtests 1 and 4, Oller (1981:483) argues that the mental processes required in subtest 1 to internally represent the dashes function

like a series of potential subjects which may be associated with predicates in a propositional manner. For instance, a logical predicate for the second dash in the series is that it is longer than the first.

In subtest 4 there are: firstly, the potential subjects “dot”, “circle”, “square”; secondly, a set of implicit predicates “inside”, “outside”; and thirdly, a set of superordinate operators “not” and “and”. One has to discover the logical proposition “dot inside circle and outside square”. This involves the testing of hypotheses that results in a mapping of the deep (abstract) propositional form into the visual elements. Oller’s (1981, p.487) point is that it is impossible to explain the mental processes revealed in the CFIT in non-propositional terms, i.e. these mental processes cannot be explained without the “deep propositional reasoning system”, i.e. without deep language.


There is no doubt that intelligence plays a crucial role in reasoning, and, accordingly, also in learning. Intelligence is one of those notions that are endlessly being bandied like a “theoretical shuttlecock” (Ryle 1949:12) from one discipline to another. In this section, I examine Oller’s (1980) “language as intelligence”, which is another term he uses to describe “deep language”, “language of thought” and the “deep propositional reasoning system”.

Some confusion could arise between Oller’s “language as intelligence” and intelligence. For example, Boyle (1987:278) rejects the view (which he describes, incorrectly, as Oller’s view) that language proficiency and “language as intelligence” are equivalent concepts and tap the same abilities. Owing to Oller’s unclear explanation of his “language as intelligence” (reasons are provided shortly), it is understandable that confusions may arise between “intelligence” and “language as intelligence”, but as Oller has made clear on several occasions, his “language as intelligence” is something deeper (deep language, LOT) than language proficiency (natural language). In the attempt to untangle the links between language proficiency, intelligence and “language as intelligence”, I shall argue that Oller’s “language as intelligence” is not identical to either “intelligence” (a distinction which Oller does not make very clear) or to “language proficiency” (a distinction which Oller does make clear, in spite of what his undiscerning critics say, e.g. Carroll, 1983a, 1983b, Boyle, 1987). (See Oller, 1991 for a restatement of his earlier views (1978, 1981, 1983a) concerning the distinction between language proficiency and “language as intelligence”).

Oller uses a non-verbal intelligence test (Cattell’s Culture Fair Intelligence Test – CFIT) in order to argue the case for the existence of deep language, but nowhere does he tackle the question of whether the CFIT tests quickwittedness, cleverness, perceptiveness, which are what I.Q. tests are supposed to measure. For some reason, Oller does not mention I.Q. No doubt, there would be quite a lot of opposition (heat?) if he had referred to Cattell’s CFIT as an I.Q. test, and a universal one at that, which is exactly what Cattell (1973) intended his CFIT to be. Culture-fair intelligence tests are regarded by many contemporary (cross-cultural) psychologists and others at best as primitive psychology, at worst as a “Western”, imperialist and elitist device for the suppression of disadvantaged cultures. Oller would possibly be accused of trying to resurrect these tests to serve some nefarious purpose disguised as philosophical speculation.

I suggest a few reasons for the rejection of the CFIT (and intelligence tests in general) by some psychologists and educationists:

1. The moral revulsion at the racial tensions that such measurements have occasioned.

2. Political pressure from egalitarian circles such as marxism and pseudo-democratic movements that discourages investigations into the relationship between intelligence and heredity (Pearson, 1991). Nowadays, the tendency in these circles is to attribute learning problems solely to causes such as the clash of cultures or political/educational oppression (e.g. Bantu Education in South Africa) or economic exploitation (e.g. capitalism).

3. The opposition of many psychologists to the purportedly simplistic (yet statistically complex) measurements of intelligence represented by an intelligence quotient (IQ).

4. Some authors such as Wober (1975:58) maintain that there are cultures that have difficulty with the general recognition of shapes. If Wober is right then these visual symbols (as they appear in Cattell’s tests) might not be culture-fair, and accordingly, an intelligence test such as Cattell’s CFIT should not be considered as a universal test of IQ. However, according to Jensen (1973:188), the shapes used in Cattell’s CFIT are universally applicable and the concepts tested are common to all societies. Oller (1981:483) maintains that although the CFIT is a visual test, it may nevertheless be argued that the mental processes required to do the CFIT are universal. Thus Oller agrees with Jensen and Cattell on the validity of the CFIT.

There is no sense in avoiding the I.Q. question because I.Q. tests are the best predictors of academic ability, which subsumes first and second language proficiency (specifically CALP) and general academic subjects (Cummins and Swain, 1986:50; Locurto’s (1991:160), Itzkoff, 1991; Pearson, 1991:111). (In Cummins and Swain is found the assertion that I.Q. tests are good predictors of second language CALP. More about CALP after the next section.


At this stage, I need to introduce a distinction between “thought” and “intelligence” (which may initially seem odd and even unnatural, but I ask the reader’s forbearance). Consider the following distinction between thought and intelligence proposed by Bohm, a quantum physicist. First, thought (Bohm, 1983:50):

Thought, considered in its movement of becoming (and not merely in its content of relatively well-defined images and ideas) is indeed the process in which knowledge has its actual concrete existence.

What is the process of thought? Thought, is, in essence, the active response of memory in every phase of life. We include in thought the intellectual, emotional, sensuous, muscular and physical responses of memory. These are all aspects of one indissoluble process. To treat them separately makes for fragmentation an confusion. All these are one process of response of memory to each actual situation, which response in turn leads to a further contribution to memory, thus conditioning the next memory.

And intelligence (Bohm, 1983:51):

The perception of whether or not any particular thoughts are relevant or fitting requires the operation of an energy that is not mechanical, an energy that we shall call intelligence. This latter is able to perceive a new order or a new structure, that is not just a modification of what is already known or present in memory…What is involved [in intelligence] is perception through the mind of abstract orders and relationships such as such as identity and difference, cause and effect, etc.

(These new orders and relationships do not have to be new to the world, but only new to the person’s mind).

For Bohm, thought is by definition conditioned, while intelligence is by definition unconditioned. But how can intelligence be unconditioned by the brainmind which is ostensibly its seat (source?) of energy? Bohm remarks in this regard:

At this point, however, one may ask: `How can one know that such an unconditioned response [i.e. an intelligent perception] is at all possible? This is a vast question, which cannot be discussed fully here. However, it can be pointed out here that at least implicitly everybody does in fact accept the notion that intelligence is not conditioned (and indeed, that one cannot consistently do otherwise).

But, if intelligence is unconditioned (by heredity and environment), it would mean that intelligence operates independently of heredity and environment. Bohm (1983:51-52) argues that if intelligence were conditioned by either heredity or environment (or both), it would then follow that the statement “intelligence is the product of heredity or/and environment” is merely the product of heredity or/and environment. This would mean that meaning is the product of heredity or/and environment, and therefore meaning would be nothing more than the conditioned “spouting forth of word patterns” (Bohm, 1983:51-52). This would reduce intelligence – and I suppose all efforts to understand the world – to conditioned “thought”.

I don’t think it is possible to deny that intelligence is to some extent conditioned by both heredity and environment. How this works is the big question, which perhaps quantum physicists (like Bohm) will be able to answer (Bohm, 1983, 1987). Or are we expecting too much from science and too little from other sources of knowledge such as theology or zen?


I now examine more closely the relationship between Oller’s “language as intelligence” as revealed by Cattell’s CFIT discussed above and language proficiency.

The kind of hypothesising revealed in the CFIT (which implies the rejection of some and the acceptance of other hypotheses) plays an important role in language proficiency (i.e. language comprehension and production). It is this kind of hypothesising that is the basis of Oller’s (1979:34) “pragmatic expectancy grammar”, which is the psychological source of “pragmatic language” (i.e. language in use).

Consider the following list of pragmatic language demands required to understand a simple story (Schank, 1982:15), all of which involve hypothesising:

1. Access and utilise raw facts.

2. Recognise stereotyped situations.

3. Make simple inferences.

4. Establish causal connections.

5. Track people’s goals.

6. Predict and generate plans.

7. Recognize thematic relationships between individuals and society.

8. Employ beliefs about the world.

Oller does not explicitly mention the pragmatic demands listed above, but I do think that he would accept them as valid “pragmatic” demands. For Oller, it is “language as intelligence” (i.e. deep language) that undergirds the abilities required to fulfil these demands listed above. In other words, deep language stands under, i.e. is the sub-stance of language proficiency.

Oller makes a distinction within language proficiency between BICS and the more sophisticated ability of CALP. (This distinction originated in Cummins, 1980, 1984). However, it still remains unclear how “intelligence” (meaning, quickwitted, perceptive) is accommodated into his language as intelligence and into the BICS-CALP distinction.

Cummins (1980, 1984) divides language proficiency into the two categories of Basic Interpersonal and Communicative Skills (BICS) and Cognitive and Academic Language Proficiency (CALP). Although it is true that BICS is the foundation of CALP and that all healthy humans beingsautomatically “acquire” BICS in their mother tongue, it does not follow that all human beings are capable of “learning” the level of CALP that is required for academic study. According to Cummins (1984), the BICS-CALP distinction is not a distinction between “communicative” language and “cognitive and academic” language [my underlining], because BICS and CALP involve both dimensions. Although I agree that both BICS and CALP are communicative as well as cognitive, I prefer to exclude the “academic” component from BICS. The reason is that although both BICS and CALP use language to think about language, CALP does this in a far more cognitively demanding way, which makes a CALP task more like an academic task, and a BICS task more like a non-academic task.

It is important to note that significant differences in intelligence (in Bohm’s sense) do not have a significant effect on the acquisition of BICS, because, as Chomksy (1967:4) points out, significant differences in intelligence (in Bohm’s sense) have only a small effect on the development of BICS. However, it is also important to note that intelligence does indeed play a significant role in the effective development of CALP.

Oller states that the major teaching problem could be a “problem of teaching language in the deep and general sense of the term” (Oller 1981:489-490). If Oller’s “language in the deep and general sense of the term” is the same notion as his “language as intelligence”, which seems to be the case, it is logical to infer that “language as intelligence” is not only “deep” language/intelligence but also “general” language/ intelligence.

The use of the term “general” in this context is confusing, because “general” intelligence in humans can refer to what Oller (1983b) – following Spearman – refers to as “g”, which is closely related to the “fluid” intelligence of Cattell (1973), where the latter refers to intelligence that is unaffected by culture or learning, and which is tested by the CFIT. Yet there is a significant difference between the lower order (what I call “general”) intelligence that all humans possess and higher order “intelligence” (as defined by Bohm above).

Regarding this confusion between these two kinds of intelligence, I focus on the (higher order) intelligence component of Oller’s “language as intelligence” and relate it to the mental processes used in the CFIT described above: If Oller’s deep intelligence is equivalent to general intelligence, then it follows that general intelligence would also be equivalent to the “deep propositional reasoning processes”, i.e. Oller’s “language as intelligence”. But, I would think that it is mostly higher order intelligence and not general intelligence that is required to grasp the orders and relationships in the CFIT, which Oller uses to illustrate the “deep propositional reasoning processes”.

To bring BICS and CALP back into the picture, Oller’s “language as intelligence” seems to subsume 1) lower order (i.e. general) intelligence (which Chomsky relates to the controversial “Language Acquisition Device” – LAD) that enables all healthy human beings to learn basic language (BICS), and 2) higher order intelligence, which is required to learn CALP. (I consider general intelligence to be belong to Bohm’s conditioned “thought”).

The salient issue is that although healthy human beings have the (lower order) general intelligence to “acquire” BICS (in their mother tongue) not all human beings have the (higher order) intelligence i.e. academic potential to “learn” CALP.

One of the major problems of many learners who enter higher primary and lower secondary school, where a second language is the medium of instruction, is that they have gained neither the necessary knowledge nor developed the necessary skills through their mother tongue to learn anything academic – whether it be (CALP in) a second language or some other subject. The fact is that second language CALP cannot be separated from first language CALP, nor can either of these be separated from proficiency in the “content” subjects, e.g. science, history.

In order to attain CALP in a first language, e.g. Chinese, one must first know BICS in Chinese. However, if a Chinese speaker wants to develop CALP in a second language, e.g. English (ESL), it is not a necessary prerequisite to develop BICS in ESL, because the attainment of a reasonable standard of BICS in ESL often only occurs after the attainment of a reasonable standard of CALP in ESL (as I have experienced with Chinese immigrants in my community). In these circumstances, CALP in a second language is developed mostly through the modes of reading and writing.


With the above distinctions in mind, consider Oller’s hierarchy of learning. Oller (1983a:355) distinguishes three levels of learning as indicated in Figure 2:

Figure 2. Oller’s hierarchy of learning

Oller’s “skills” of listening, etc. in Figure 2 refer to the general skills of listening, speaking, reading and writing as well as the deeper cognitive skills that are learnt through these four modes of language.

So far, I have emphasised the central role of deep language in learning, but it is important to keep in mind that in the learning process the development of deep language depends (partly?, mostly?, entirely?) on the development of natural language. A dialectical and internal relationship exists between the two levels, in fact, between all the levels in the hierarchy of learning. What occurs is not merely a top-down bottom-up exchange between the three levels, but a far more complex transaction between biomental mechanisms, intelligence, linguistic knowledge, world knowledge and skills.

The fact that Oller does not include other important factors in his hierarchy of learning does not mean that he is unaware of their importance. Some of these factors are motivation, acculturation, and mother tongue interference. Besides these aforementioned factors, there are also economic and sociopolitical factors that may play a significant role in learning, e.g. the unequal distribution of resources, or strikes and boycotts (which have been so rampant in South African education in recent times). The point is that many factors may influence learning, but none of them (including natural language), in my view, is as fundamental as the attribute of deep language. And this is what Oller’s hierarchy attempts to show.

A possible defect in Oller’s hierarchy is that the inclusive term of deep language does not distinguish between intelligence, in Oller’s general sense of a “deep propositional reasoning system”, and intelligence, in the specific sense of finding new order, and so forth.

A possible defect in Oller’s hierarchy is that the inclusive term of deep language does not distinguish between intelligence, in Oller’s general sense of a “deep propositional reasoning system”, and intelligence, in the specific sense of finding new order, and so forth.

Bohm’s distinction between (“conditioned”) thought and   (“unconditioned”) “ntelligence  comes in useful here. For example, early language learning is a product of conditioned thought. Chomsky shows very clearly that child language acquisition is conditioned, i.e. automatic, i.e. it is built into the hardwiring of the brain, into what he calls the Language Acquisition Device (LAD). Whether LAD is a distinct “organ” is another matter, which won’t occupy us here. But all language cannot be conditioned; if it was, meaning would be nothing more than the conditioned “spouting forth of word patterns” (Bohm, 1983:51-52), as discussed above. Intelligence (in Bohm’s sense) enables us to make deductions and inductions (and “abductions”, as discussed in the works of C. S. Pierce; see Nagel, 1959; Oller, 1991:6).

For example, the existence of LOT is partly derived from (induction) the empirical observation that children learn language (grammar and lexis) automatically, and partly from the reasoning that in order for a child to know how to mean (which involves both thinking and expressing “in” a linguistic system), the child must already have some innate mechanism of knowing how to mean. This innate mechanism, as it has been argued in this article, is LOT (deep language). The concept of LOT is arrived at not only through a process of inductive reasoning, but also through hypothesising about how language can exist at all. The epistemological journey (i.e. process of acquiring knowledge) takes us from child language learning (empirical data) and philosophical speculation (rational data) to LOT. Much “intelligent” learning is based on these two directions of enquiry. The ontological discovery at the end of the journey is that LOT existsprior to (natural) language. To oversimplify: the empirical data is already given; it is left to theorists to take (to) the theory that best suits their view of the world – a rationalist view, a behaviourist view, or a what have you . According to Dewey, ideas are not “givens” but “takens”. Point taken.


This article dealt with the notion of deep language (“language as intelligence”) and showed some of its connections to thought, intelligence, language proficiency and learning. The (unusual) distinction between “thought” and “intelligence” was used to clarify Oller’s confusing notion of “language as intelligence”. It was also argued that Basic Interpersonal and Communicative Skills (BICS) requires lower order (general) intelligence, while Cognitive and Academic Language proficiency (CALP) requires higher order intelligence. An understanding of the distinction between BICS and CALP, and their relationship to intelligence is indispensable in understanding the process of (academic) learning.


Bloch, M. 1991. Language, anthropology and cognitive science. Man, 26, 183-198.

Bohm, D. 1983. Wholeness and the implicate order. London: Ark Paperbacks.

Bohm, D. 1987. Unfolding meaning: A weekend of dialogue with David Bohm. London: Ark Paperbacks

Boyd, R. and Richerson, P.J. 1985. Culture and the evolutionary process. Chicago. University of Chicago Press.

Boyle, J.P. 1987. Intelligence, reasoning, and language proficiency. Modern Language Journal, 71, 277-288.

Carroll, J.B. 1964. Language and thought. Englewood-Cliffs, New Jersey: Prentice Hall.

Carroll, J.B. 1983a. Psychometric theory and language testing. In J.W. Oller, Jr. Issues in language testing research. Rowley, Massachusetts: Newbury House.

Carroll, J.B. Studying individual differences in cognitive abilities: through and beyond factor analysis. In R. Dillon, and R. Schneck. (eds.). Individual differences in cognition. New York: Academic.

Cattell, R.B. 1973. Measuring intelligence with culture-fair tests. Institute of Personality and Ability Testing.

Chomsky, N. 1967. Recent contributions to the theory of innate ideas. Synthese, 17, 2-11.

Chomsky, N. 1972. Language and mind. New York: Harcourt Brace Jovanovich.

Cummins, J. 1980. The cross-lingual dimensions of language proficiency:Implications for bilingual education and the optimal age issue. TESOL Quarterly, 14, 175-87.

Cummins, J. 1984. Wanted: A theoretical framework for relating language proficiency to academic achievement among bilingual students. In C. Rivera, (ed.). Language proficiency and academic achievement. Multilingual Matters 10. Clevedon: Multilingual Matters Ltd.

Cummins, J. and Swain, M. Bilingualism in education. 1986. London: Longman.

Danesi, M. 1994. The neuroscientific perspective in second language acquisition research: a critical synopsis. International Review of Applied Linguistics, 32(3):201-228.

Daniels, H., Lucas, N., Totterdell, M. and Fomina, O. Humanisation in Russian education: A transition between state determinism and individualism. Educational Studies, 211):29-39.

Devitt, M., and Sterelny, K. 1987. Language and reality: An introduction to the philosophy of language. Basil Blackwell

Fodor, J.A. 1986. Psychosemantics. Cambridge, Massachusetts: The MIT Press.

Fodor, J.A., Bever, T. and Garrett, M.E. 1974. The psychology of language: An introduction to psycholinguistics and generative grammar. New York: McGraw-Hill.

Gee. J.1990. Social linguistics and literacies: Ideology in Discourses. London. The Falmer Press.

Gregory, R.L. 1981. Mind in science: A history of explanations in the psychology of physics. Cambridge. Cambridge University Press.

Groddeck, G. 1977. The meaning of illness. London: Hogarth Press.

Halliday. M.A.K. 1977. Learning how to mean. The Netherlands: Elsevier, North Holland.

Hayakawa, S.I. and Hayakawa, A.R. 1990. Language in thought and action (5th Edition). New York: Harcourt Brace Jovanovich.

Itzkoff, S.W. 1991. Human intelligence and national power. New York: Peter Lang.

Jensen, A.R. 1973. Educational differences. London: Methuen.

Krashen, S. 1981. Second language acquisition and second language learning. Oxford: Pergamon Press.

Lenneberg, E.H. 1967. Biological foundations of language. New York: John Wiley and Sons, Inc.

Locurto, C. 1991. Sense and nonsense about IQ:The case for uniqueness. New York: Praeger.

Loewer, B. and Rey, G. 1992. Meaning in mind:Fodor and his critics. Oxford:Blackwell.

Nagel, E. 1959. Charles Sanders Pierce. a prodigious but little known American philosopher. Scientific American, 200:185-192.

Nolan, R. 1993.Cognitive practices:Human language and human knowledge. Oxford: Blackwell.

Oller, J.W., Jr. and Perkins, K. 1978. Language in education: Testing the tests. Rowley, Massachusetts: Newbury House.

Oller, J.W., Jr. 1979. Language tests at school. London: Longman.

Oller, J.W., Jr. 1981. Language as intelligence. Language Learning, 31, 465-492.

Oller, J.W., Jr. 1983a. Issues in language testing research. Rowley, Massachusetts: Newbury Publishers.

Oller, J.W. (Jr.). 1983b. “g”, “What is it? In: Hughes, A. and Porter, D. (eds.). Current developments in language testing. London: Academic Press.

Oller, J.W. Jr. 1991. Language and bilingualism: More tests of tests. London and Toronto: Associated University Presses.

Paradis, M. 1991. Language lateralization in bilinguals. Brain and language, 39, 576-586

Pearson, R. 1991. Race, intelligence and bias in academe. Washington, D.C: Scott-Townsend Publishers.

Perecman, F. 1989. “Language processing in the bilingual: Evidence from language mixing. In: K. Hillenstain and L.K. Obler (eds.). Bilingualism across the lifespan. Cambridge. Cambridge University Press.

Pribham, K. 1971. Languages of the brain. Englewood Cliffs, New Jersey: Prentice-Hall

Ryle, G. 1949. The concept of mind. Hutchinson.

Schank, R.C. 1982. Reading and understanding: Teaching from the perspective of artificial intelligence. Hillsdale, New Jersey: Lawrence Erlbaum Associates.

Vygotsky, L.S. 1962. Thought and language. Cambridge, Massachusetts: The MIT Press.

Vygotsky L.S. 1978 Mind in society: The development of higher psychological processes. Edited by M. Cole, V. John-Steiner, Scribner, S., and E. Souberman, Cambridge, Massachusetts: Harvard University Press.

Wober, M. 1975.Psychology in Africa. London: International African Institute.

Woese, C.R. 1967. The genetic code: the molecular basis for genetic expression. New York: Harper and Row.

Young, D. 1990. How long will applied linguistics survive? or Are transformations still possible? In K. Chick (Ed.). Searching for relevance: Contextual issues in applied linguistics. South African Applied Linguistics Association.

Young, D. 1993.NEPI: Language process or product? Paper presented at the 12th Annual South African Association of Applied Linguistics (SAALA) Conference, University of Port Elizabeth, 28-30 June.

One response to “Deep Language, Intelligence and Language Proficiency in Learning

  1. Pingback: Cognition and Language Proficiency « Grammargraph

Leave a Reply

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

You are commenting using your 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: