Most of us take our ability to communicate for granted. We think of something to say, open our mouths, and the words come out. Although it may appear to be a simple thing, the communication process is actually very complicated. The complexity of human communication becomes especially apparent when we consider the diversity that exists among individuals for whom communication is difficult.
The study of communication and its disorders has occupied scientists and philosophers for many years. The ancient Greeks, for example, described disorders of speech and treatments. Over the years, many scientists have attempted to develop classification schemes to describe disorders of communication, and work continues in this area today. In this paper, I summarize several basic concepts related to language and its assessment by speech-language pathologists (trained professionals who study, assess, and treat disorders of communication).
Language vs. Speech
The terms speech and language are related, but refer to different aspects of the communication process. We’ll begin our discussion by considering the term ‘language’ first. Language has been defined by Crystal (1987) as: The systematic, conventional use of sounds, signs, or written symbols in a human society for communication and self-expression (p. 424).
Language is a complex, often arbitrary, rule-governed system. It is complex in that we can identify subsystems of language (see below). It is often arbitrary in that words and language rules are likely to differ from one language community (such as American English) to another (such as French). In other words, there is nothing about four legged animals who bark that makes us use the word dog to describe them; we could use the word chien just as easily. Finally, language is rule-governed in that speakers of a language agree that some constructions are permitted (for example, John likes to read books), whereas other constructions are prohibited as ungrammatical (for example, Read John books to likes).
Linguists (scientists who seek to describe and explain language) often differentiate among five subsystems of language:
Pragmatics: Rules that govern language use.
Pragmatic abilities include language function (for example, what do we hope to achieve through communication), directness (give me some pie is more direct than that pie looks good), appropriateness, as well as interpersonal factors such as assumptions regarding shared knowledge, knowing when to introduce or change topics, etc.
Semantics: Rules that govern language content.
The Semantic subsystem is concerned with the meaning of words (vocabulary) as well as the role that words play in communication. For example, consider the following sentences:
Pat hugged the doll.
The doll was hugged by Pat.
In both sentences, Pat was the “hugger” (the agent) and the doll was the “huggee” (the object receiving the force of the action ‘hug’).
Syntax: Rules that govern formulation of grammatical sentences, especially ordering of words.
Syntactic rules dictate the structure of phrases and how those phrases are combined into sentences. In addition, syntax allows us to rearrange words to make new sentences. For example, the words in the declarative sentence It is raining. can be rearranged to ask the question, Is it raining?
Morphology: Rules associated with the formation of words.
Words can be broken down into smaller units called morphemes. The word cats, for example, is made of two morphemes: cat meaning feline and -s meaning more than one. Morphology and syntax comprise what many of us think of as grammar.
Phonology: Rules associated with the sound system of a language.
The rules of phonology determine what sounds are used by a speaker of a language, when they are used, and how they can be combined. Also, phonological rules describe sound changes. For example, we pronounce the word electric with a [k] sound at the end. If we add the suffix -ity to form electricity, then the pronunciation changes and the [k] sound is lost and an [s] is pronounced instead.
From this description, we can see that language is a complex, abstract concept. Language can be written, spoken, or signed (as in American Sign Language). The spoken form of language is speech. Speech parameters include production of voice, movement of the oral structures to produce speech sounds (articulation), and smoothness of speech production (fluency).
Receptive vs. Expressive Language
As communicators, we use our language abilities to make sense of what others say to us and to express our needs, thoughts, and emotions in ways that others will understand. It is sometimes useful to compare performance on tasks that assess an individual’s ability to understand language (receptive language) with performance on tasks that emphasize language formulation and production (expressive language). In a subset of individuals with communication disorders, the difference between receptive and expressive language performance may be significant, with the expressive language score typically being the lower of the two.
Quantitative vs. Qualitative Assessment of Language
There are two basic approaches to measuring language behavior: quantitative and qualitative. Quantitative approaches use numbers (scores) to reflect standing relative to other individuals of the same age (or grade, in the case of children who have been ‘held back’ or ‘promoted’ a grade). Qualitative approaches de-emphasize the use of numbers and attempt to describe the individual’s language. Standardized tests (such as the Test of Language Development – Primary:3 [TOLD-P:3] or the Clinical Evaluation of Language Fundamentals – 3 [CELF-3]) are good examples of the quantitative approach to language measurement. Analysis of a spontaneous sample of language elicited under naturalistic conditions would be an example of a qualitative approach.
In the literature, many researchers have taken sides. That is, they say that one form of evaluation is good and the other is bad. My view is that quantitative and qualitative assessment are complementary means of evaluation and together they can give us a clearer picture of a child’s strengths and needs than either approach can achieve alone.
Quantitative approaches tend to be time-efficient, which is one reason they are so widely used. They involve standard procedures. That is, all individuals who administer a test such as the TOLD-P:3 will use the same pictures, give the directions the same way, and use the same scoring methods. Such tests are typically norm-referenced, and will give us a very general idea as to how the child’s communication skills compare with those of other children the same age.
Some standardized tests are made up of subtests that seek to provide information regarding the different language channels or subsystems described above. These test are designed to allow us to formulate or test certain hypotheses. Part of being human is that we are not equally good at all things. For example, I’m pretty good at science, but in terms of physical ability, I’m a total klutz! A well standardized test can give us an idea as to what areas function as relative strengths and which are relatively weak. For example, we might compare performance on tasks emphasizing receptive language with those that emphasize expressive language. Likewise, some standardized tests facilitate comparisons among the language subsystems: semantics, syntax, phonology, etc. A well standardized test provides information regarding reliability of measurement, which affects our ability to interpret the test results. The more reliable the test, the greater the likelihood that we would get similar scores if we were to re-test the child. In my opinion, one of the strongest attributes of a well standardized test is that it will tell us how bad it is (in terms of reliability and validity information).
Drawbacks of quantitative assessment? Well, there are many. Quantitative methods tend to be rather superficial. Even a well standardized test, such as the TOLD-P:3 or the CELF-3, uses a rather small sample of words to assess, for example, vocabulary skills. The sample in many cases is too brief to provide an adequate picture of the child’s abilities. Similarly, there are many language skills (e.g., pragmatics) that are not assessed at all by tests such as these. To complicate matters, published tests vary in quality. Unfortunately, not all tests can be described as well standardized. It is important to note that abilities beyond language (for example, hearing, vision, attending, compliance) may be contributing factors when an individual earns a low score on a norm-referenced language test.
Qualitative approaches are not standardized. Interacting naturally with a child and documenting his/her language production provides a sample that is likely to be closer to “real life” than a standardized test. We can analyze the sample in different ways to look at different types of skills, thus the sample is more flexible than a standardized test. The obtained sample is typically longer than the sample obtained from a standardized test, and analysis of the sample is often useful for the purposes of planning an intervention program.
Transcribing and analyzing the sample is often time-consuming. In addition, the validity of the sample is affected by the skills of the individual collecting the sample, the mood of the child, the rapport that exists between the clinician and child, etc. (Many of these threats are also true of standardized testing). Another negative aspect of qualitative approaches is the difficulty associated with evaluating reliability. The only way to know my sample is reliable is to take a second sample and analyze it and compare the two. In clinical sites, this is seldom possible.
A Personal Perspective
When I seek to assess a child’s language skills, I employ both quantitative and qualitative procedures. I analyze the child’s performance on standardized tests before I analyze the spontaneous sample of language. High scores on a standardized test may help us to rule-out certain types of problems. Low scores beg for an explanation and prove to be a challenge to interpret. A low score could be related to low language ability (it could also be related to other factors such as attending, hearing, vision, compliance, etc.) The standardized test allows me to develop hypotheses, which can then be tested using qualitative approaches (and/or additional follow-up testing). Qualitative approaches are less time-consuming and more likely to provide insight when I have an idea as to what I am looking for. As a diagnostician, I am like a detective who tries to make sense of clues obtained through quantitative and qualitative procedures. I strive to identify the explanation that best accounts for the patterns I observe.
We must always remember that the diagnosis of communication disorders is complicated because such impairments may affect any or all of the five linguistic subsystems and/or the various speech parameters. It is important to understand, too, that language and speech are interdependent in the sense that certain language problems will result in problems that appear related to speech output. For example, some children do not use the plural -s suffix, and will say things like two cat rather than two cats. It may be difficult to determine whether the error is associated with a language skill (morphology) or is due to a speech skill (articulation of [s]) or both. Likewise, a child with a severe speech problem may omit grammatical words such as the, is, etc. because the act of producing speech is laborious. For example, the child may say, cat run rather than the cat is running. It is often difficult to determine whether the child’s production is due to a speech problem, a language problem (morphology, syntax), or both.