Salience and difficulty

Discussions of ways of improving stimulus control over a target behavior sometimes refer to augmenting the “salience” of antecedent stimuli.  Such uses of this term are formalized in applied behavior analysis textbooks.  For example, Cooper, Heron, & Heward (2007) define salience as “the prominence of the stimulus in the person’s environment” and add that “Some stimuli have more salience than others depending on the sensory capabilities of the individual, the past history of reinforcement, and the context of the environment” (p. 401).  Miltenberger (2008) indicates that a stimulus is salient “when it is intense or easily detected by the individual” (p. 593).  He also uses “noticeable” and “conspicuous” as synonyms.  Sulzer-Azaroff & Mayer (1991) discuss emphasizing or enhancing relevant stimulus properties when essential features are not “sufficiently salient or obvious to the learner” (p. 270, italics in the original).  These references are consistent with colloquial usage, which equates salience with something being striking, conspicuous, noticeable, prominent, and so forth.

What is salience?

The problem with these kinds of quasi-technical references in our professional dialect to the salience of antecedent stimuli is that they come uncomfortably close to implying that salience is a physical dimension of a stimulus that can be independently identified, measured, and manipulated.  Catania (2013) succinctly points out the problem: “Stimulus properties to which an organism is likely to respond are sometimes called salient, but salience isn’t a property of a stimulus; it’s actually a property of the organism’s behavior with respect to that stimulus” (p. 140, italics in the original).

The fact that salience isn’t an aspect of a stimulus is evident in the absence of any means by which such features might be consistently defined and measured.  Dimensional quantities of physical events (e.g., mass or wavelength) have standard definitions in terms that are independent of the event being measured (Johnston & Pennypacker, 2009).  In contrast, a stimulus dimension of salience cannot be defined independently of any particular stimulus or even independently of behavior.  The only way one can define and measure the salience of a stimulus is in terms of changes in behavior when particular stimulus features are varied.  For instance, a complex of stimuli, a specific stimulus, or particular feature of a stimulus might be described as salient if responding was more likely in its presence than its absence.  The fact that salience cannot be defined independently of changes in behavior clarifies that it is not a dimensional feature of a stimulus.  Instead of referring to manipulating the salience of a stimulus, it is more appropriate to discuss changes in physical features of a stimulus that might control responding more or less effectively.

And difficulty?

This same argument applies to discussions of the “difficulty” of tasks or test items.  Changing the difficulty of a task by manipulating its features is often viewed as a means of making responding more or less likely (e.g., Weeks & Gaylord-Ross, 1981).  As with salience, however, difficulty isn’t a physical feature or dimension of a stimulus.  What is changed by manipulating physical features of task stimuli is not difficulty, but merely the stimuli themselves.  The only way of describing these changes as making the task more or less difficult requires measuring changes in responding.  In other words, when we refer to difficulty, we are describing the effect of environmental events on responding, not characteristics of the events themselves.  The same stimuli comprising a task can obviously have different effects on responding from one individual to another or even for a single individual under different circumstances.

Perhaps we tend to talk as if salience and difficulty are physical aspects of stimuli because they serve as nouns in our grammar.  What we are tacting, however, is behavior, not the environmental circumstances in which responding occurs.

References

Catania, A. C. (2013). Learning, 5^th edition. Cornwall-On-Hudson, NY: Sloan Publishing.

Cooper, J. O., Heron, T. E. & Heward, W. L. (2007). Applied behavior analysis. Upper Saddle River, NJ: Pearson Prentice Hall.

Johnston, J. M.,& Pennypacker, H. S. (2009).  Strategies and tactics of behavioral  research. NY: Routledge.

Miltenberger, R. C. (2008).  Behavior modification, 4^th edition. Belmost, CA: Thomson Wadsworth.

Sulzer-Azaroff, B. & Mayer, G. R. (1991). Behavior analysis for lasting change. Orlando, FL: Holt, Rinheart and Winston.

Weeks, M. & Gaylord-Ross, R. (1981). Task difficulty and aberrant behavior in severely handicapped students. Journal of applied behavior analysis, 14, 449-463.