Neal Miller

Explaining language: A behavioral critique of Skinner’s analysis of verbal behavior.

In Verbal Behavior (1957), Skinner attempted to offer a functional account of human language and made a point of contrasting his approach with the more traditional accounts available at the time. Rather than focus on the structure or mechanics of language (formal aspects of language), Skinner attempted to identify the conditions that gave rise to those behaviors. Although Skinner’s analysis of verbal behavior has been influential, particularly in treating language deficits for individuals with autism, there are conceptual problems with the way he defined and categorized verbal behavior. In this paper we argue that Skinner’s analysis is in fact largely based on formal aspects (rather than functional) and that this has both created confusion and limited the utility of the analysis. Specifically, we argue that Skinner’s formal account makes it difficult to distinguish verbal from nonverbal behavior and to distinguish the various types of verbal responses from one another. We then summarize and respond to some of the contemporary defenses and criticisms of Skinner’s analysis of verbal behavior. Finally, we argue that although Skinner’s analysis has had some practical utility, the conceptual benefits are questionable. S C I E N T I F I C A B S T R A C T Skinner (1957) offered a behavioral account of language in his book Verbal Behavior. Compared to more traditional, structural accounts of language at the time, Skinner’s analysis attempted to identify the variables that control and maintain verbal behavior. This behavioral account of language has proven useful, especially in the area of treating language deficits for individuals with autism. However, there exist conceptual problems with Skinner’s analysis. The definition of verbal behavior and subsequent taxonomy of verbal operants (or units) is based largely on formal properties. We suggest that these formal elements of Skinner’s analysis result in arbitrary distinctions that emphasize the form of behavior or stimuli. This is problematic, at a conceptual level, as distinguishing between verbal and nonverbal behavior and distinguishing between the different verbal operants necessitates not only a functional account of stimuli but identification of the source and form of stimuli. In this paper, we examine some of the contemporary defenses and criticisms of Skinner’s analysis. Lastly, we conclude that although Skinner’s analysis of language has had practical utility, the conceptual benefits are limited.

Functional analysis results across a known and unknown assessor

ABSTRACT Functional analyses determine behavior function by examining response patterns across various contingency conditions. Typically, functional analyses are conducted by a trained clinician with whom the participant is often unfamiliar. This unfamiliarity might influence the outcome of the analysis, potentially leading to misidentification of function. This study examined the responding of two participants during functional analyses conducted by a known and unknown assessor. The influence of familiarity was initially observed across both participants; however, with extended exposure to the functional analysis responding became similar across assessors for both participants. Importantly, different conclusions regarding function may be drawn at various points during the analysis. These findings are discussed with regard to the influence of participant–assessor history and recommendations are provided regarding conducting a functional analysis.

Assessing Writing in Elementary Schools: Moving Away From a Focus on Mechanics

ABSTRACT Many students struggle with writing; however, curriculum-based measures (CBM) of writing often use assessment criteria that focus primarily on mechanics. When academic development is assessed in this way, more complex aspects of a student’s writing, such as the expression and development of ideas, may be neglected. The current study was a preliminary analysis of viability of the words per thought unit (T-unit) as an alternative to traditional approaches to writing assessment. Participants included 167 children enrolled in the fourth and fifth grade, ages 9–11, in a rural southeastern U.S. school district. Writing samples were collected in three probes across the school year (fall, winter, and spring) and analyzed by using a variety of measures. Words per T-unit were positively linearly correlated with other measures of literary proficiency. A repeated measures analysis of variance (ANOVA) assessed growth in writing across time and grade. Results indicated a statistically significant time effect within subjects for words per T-unit across grades. Thus, the analysis of words per T-unit yielded promising results as an efficient means of measuring writing skills independently of mechanics.