The Linguistic Era 1965-1975

The new approach to viewing children's disordered language as a linguistic system did not evolve from our aphasia-in-children thrust, but rather from events outside our field in psychology and linguistics. The event which is usually seen as starting it off was Noam Chomsky's transformational linguistic theory which postulated an abstract level of language used by people to understand and produce sentences. Following Chomsky's idea of an abstract generative language system (Chomsky, 1957) psychologists, postulated rules for beginning language learners, rules that are different from adult users of the language.

Perhaps the first study to show the linguistic regularity and generative nature of children's language was the 1958 study by Jean Berko Gleason on children's productive use of morphology (Berko-Gleason, 1958). She argued from her data that what the children were learning was rules and not just responses to stimuli that were then generalized to similar stimuli.

Gleason's mentor, Roger Brown conducted and directed other pioneering studies at Harvard and MIT during this time. For example, Brown traced children's acquisition of 14 English morphemes (Brown, 1973), Brown supervised a dissertation of Ursula Bellugi on children's acquisition of negation (Bellugi, 1967); and Fraser, Bellugi and Brown published their well known study on children's imitation, comprehension and production of selected syntactic structures (Fraser, Bellugi, & Brown, 1963).

Roger Brown's core group of students went out over the next few years to do their own research in child language and to lead the way in tying research in child language to structural or transformational linguistics. They were the Johnny Appleseeds of developmental psycholinguistics.

Meanwhile, what was happening back at home in speech pathology? Three people in our discipline were translating for us the research methodology and findings of the psycholinguists: Paula Menyuk, Laura Lee, and Elizabeth Carrow. Paula Menyuk, who was watching what was going on at MIT from her perch at Boston College, incorporated transformational grammar ideas into her doctoral dissertation as early as 1961 (Menyuk, 1961). Her postdoctoral fellowship at MIT following her dissertation allowed her to develop and publish her research on both normal (Menyuk, 1963) and language disordered children (Menyuk, 1964). Her emphasis, like Chomsky's, was on syntax, and the transformational rules that serve to produce different types of sentences. Menyuk's research method was to study preschoolers' and kindergartners' acquisition of different sentence types as evidenced by their ability to imitate sentences.

Laura Lee at Northwestern University was also seeing the potential usefulness of linguistics for work with children who had language disorders In fact, Lee's interest in linguistics had predated Chomsky's revolution by a number of years. By 1966 Lee had already read widely in general semantics, and was attuned to the literature in structural linguistics, that same literature which Chomsky was weaned on (Lee, 1958). Perhaps, then, Lee was the natural person to devise the first analytic tool using linguistic categories to analyze language samples for children at the two-word stage (Lee, 1966) as well as for children whose language had already acquired simple basic sentence structure (Lee & Canter, 1971). Development of the Northwestern Syntax Screening Test by Lee also showed her awareness of the literature on developmental psycholinguistics (Lee, 1971). The test was modeled after Fraser, Bellugi and Brown's study of children's imitation, comprehension and production of syntactic forms.

The year 1971 also witnessed the publication of the Test for Auditory Comprehension of Language by Elizabeth Carrow designed to measure children's comprehension of lexical, morphological, and syntactic information (Carrow, 1971). Following this, a number of therapy programs were reported in the literature on enhancing children's comprehension of prepositions (Frisch & Schyumaker, 1974), their understanding of directions containing noun-verb combinations (Striefel, Wetherby, & Karlan, 1976), and their comprehension of other grammatical sequences (Stremel & Waryas, 1974). The programs were highly structured ones, carried out in drill fashion, using behavioral principles.

The new emphasis on syntax assessment was accompanied by the emergence of syntax teaching programs for language production. Clinicians who embraced Chomsky's transformational grammar and nativist ideas were confronted with a dilemma that what they wanted to teach children were abstract language rules?rules that were never heard or spoken, that never appeared in surface structure. Those purists who believed with Chomsky that the necessary knowledge was innate, didn't develop new clinical approaches to therapy. The only teaching methods around were those developed by behaviorists and their concern was that behaviorism treated language structure as surface regularity. The outcome of this historical dilemma was that the behaviorists developed syntax teaching programs using learning theory principles of associationism (stimulus?response drill) and reinforcement (Gray & Ryan, 1973; Stremel & Waryas, 1974) There were a few who developed syntax teaching programs which placed less emphasis on behavioral technology (Fokes, 1976; Tyack & Gottsleben, 1977), but they, too, assumed that getting the clients to imitate or practice a pattern of sentences correctly was an indicator that they had the deep structure understanding of the underlying rule.

Laura Lee again stood in the forefront of this period. Unlike those who taught morphology and syntax rules via unrelated sentences drills, Lee with Susan Mulhern, devised naturalistic conversational approaches to teaching linguistic rules. They contextualized the morphological and syntactic targets by presenting them in story formats (Lee, Koenigsknecht, & Mulhern, 1975).

During the early 70's psycholinguistic theory shifted from an emphasis on syntactic structure to semantics. Second generation psycholinguists such as Lois Bloom (Bloom, 1970) and Melissa Bowerman (Bowerman, 1976) argued convincingly that the child could use the same surface sentence to express different meanings and that we must attend to the cognitive and semantic notions underlying those surface syntactic structures. So emerged semantics teaching programs?in particular ones that focused on two word relations such as possession, location, action?patient, and agent?action (Leonard, 1975; MacDonald & Blott, 1974; Miller & Yoder, 1974). The means for teaching such mentalistic constructs were still laden with behaviorism in that the clinicians were instructed to use imitation and reinforcement in game or drill formats, with a focus on getting the child to produce the right responses.

The linguistic move had included syntax and semantics by the mid 70s, and by late 1970's yet another area was brought into the fold. This time the emphasis was on the linguistic aspects of phonology. The approach, called phonological process analysis, pioneered by linguist David Ingram in his book, Phonological Disability in Children. (Ingram, 1976). The approach differed from articulation analyses in that it introduced an abstract level of knowledge, the phonological version of syntactic deep structure.

Actually, a step toward thinking of articulation problems as coming from abstract linguistic rules preceded Ingram's work, in the cloak of distinctive feature theory. Phonemes were viewed as being made up of features, with the same feature occurring in very different phonemes. The feature approach laid the groundwork for the process approach in that different phoneme errors such as t for s, or d for z, or p for f were considered the same-in this case an error of frication. The frication feature then became the targeted structure for therapy rather than the individual phoneme (Blache, 1982; Costello & Onstine, 1976; McReynolds & Houston, 1971).

During this linguistic period we experienced the familiar theory-therapy gap. It was between new mentalistic insights and clinical methods derived from behaviorism. Clinicians incorporated feature theory into their teaching goals and carried out their lessons in behavioral theory. The therapies included discrimination, phonetic placement and imitative drills (McReynolds & Bennett, 1972;; Costello & Onstine, 1976).

But the technology of behaviorism being used to teach linguistic forms was not just a simple outgrowth of behavioral theory. Rather, it was a much more elaborate and specified technique called behavior modification. It was built from behavioral assumptions in that the lessons moved from small analytic units to larger ones, and stimuli, responses and reinforcement were the pivotal constructs around which the programs were organized. The technique went well beyond what would go on in ordinary everyday learning. For example, the behavior modification technique prescribed a schedule of reinforcement and included well-developed techniques such as imitating a stimulus, fading the stimulus, shaping the response, successively approximating the target. The object of behavior modification for teaching language was to add units to verbal repertoires and to learn the proper associations between particular stimulus conditions and the required responses (e.g.,Mowrer, 1977, Holland & Harris, 1968).

In most behavior modification programs that were dedicated to teaching language behavior, the stimulus was a verbal one, a sound or syllable or sentence, and the response was the child's imitation of that stimulus. The reinforcement was an edible, verbal or token reward given for correct responses according to a predetermined reinforcement schedule (Gray & Fygetakis, 1968; Sloane & MacAulay, 1968).

A number of programs were programmed to build logically from one stimulus-response association to another. For example, the Monterrey program for teaching grammatical constructions is a logically staged set of grammar teaching programs in which a teacher shows a child a picture and provides an associated verbal stimulus (Mark, the man says: I am going to the store) requiring the child to either imitate part or all of the stimulus (I am going to the store) or to complete it (The man says: I-Mark? Mark: I am going to the store) (Gray & Fygetakis, 1968; Gray & Ryan, 1973).

Another, more mentalistic approach to language intervention during this period between 1965 and 1975 was to focus on the language the child was hearing from caregivers. Courtney Cazden, was the first to attend to this in her dissertation of 1965 (Cazden, 1965). In that work she found that when adults consistently expanded upon what the child said, the child's language learning was enhanced. Catherine Snow continued this line of research on child-directed talk to typical language learners (Snow, 1972), soon to be followed by those who used techniques of semantic expansion with children who had language disorders (MacDonald, Blott, Gordon, Spiegel, & Hartmann, 1974). Changing the way adults talked with children has been referred to as indirect language stimulation techniques or incidental teaching. This approach came into full bloom in the work of Ayala Manolson and her colleagues at the Hanen Centre in Toronto Canada (Manolson, 1992).

In 1972 Van Riper reemerged with a new edition of his Speech Correction: Principles and Methods (Van Riper, 1972) and yet again in 1978 (Van Riper, 1978). Tellingly, the 1963 chapter entitled delayed speech was called delayed speech and language in 1972, and called delayed development of language in 1978. While his earlier versions had chapters entitled "the development of speech" or "how children learn to talk" his 1978 edition renamed it "the development of speech and language. In the 1972 edition chapters included a section on "building sentences and by 1978 it had expanded to include a section on phonological development, semantic development, as well as continuing the syntax section, calling it "learning to talk in sentences." In both 1972 and 1978 editions Van Riper lists the characteristics of brain-damaged children treating it as a syndrome causing language problems in children. He also includes sensory deprivation, mental retardation, and autism, as a nod toward those of Myklebust's differential diagnosis persuasion.

Van Riper's therapy techniques also reflect the times. On the one hand he lists operant conditioning and imitation, and on the other, modeling as self and parallel talk. In 1978 he has a section called "The linguistic approach: Discovering the hidden rules of language." He includes in the chapter indirect language stimulation techniques techniques of modeling, expanding, and extending children's language productions.

Summary, so far

In the course of our stroll thorough the about the first seventy-five years of our history from 1900 to 1975, we have identified certain milestones which have lead us to contemporary lines of thought in working with language impaired children. Our earlier foundations were based on behavioral assumptions of atomism and peripheralism. The atomistic notion was that phonemes were the building blocks or atoms of speech. The peripheralist notion was that therapy methods were based on stimulus-response associations, in this case work on stimuli meant speech sound bombarbment and auditory discrimination of sounds, and on responses meant phonetic placement for speech sound production.

With the termination of the second-world war there was a new emphasis on developing frameworks for conducting therapy with aphasic veterans. We began to borrow from the work on aphasia to develop methods for working with language-impaired children. The aphasia model of diagnosis and therapy originated in Europe and called for a distinction between speech and language. Thus came the use of terms like inner language, or mediated responses, and a set of diagnostic and therapeutic methods for working to improve the language of aphasic children. In our third period, beginning in the early 1960s, we concentrated on language as a structured system, with its own rules and its layers of syntax, semantics and phonology. Van Riper's various revisions of his influential text can be seen as a litmus test for what changes were going on in the mainstream of our profession. During this period he changed from an emphasis on auditory discrimination and phonetic placement, without accompanying language training (Van Riper, 1939), to one which distinguished language from speech (Van Riper, 1972) and finally to one with sections on syntax, semantics and phonology (Van Riper, 1978).

We now turn to a new evolution of therapies, ones that treated language as part of the social, language, and situational context. This came to be known as the pragmatics revolution, a shift in emphasis that began in the late 1970s.