Elsevier

Behavioural Processes

Volume 56, Issue 2, 1 November 2001, Pages 67-74
Behavioural Processes

Context modulation of US signal value following explicit and nonexplicit training

https://doi.org/10.1016/S0376-6357(01)00186-3Get rights and content

Abstract

Context modulation of an unconditioned stimulus (US) signal value was examined in three groups of rats. In Group A+/B−, subjects received explicit training in which a single food pellet US was followed by three additional pellets after either a short delay (in Context A) or a long delay (in Context B). Two other groups received nonexplicit training in which subjects received either the same conditions in Context A, but simple exposure to Context B (Group A+/Bo), or the same conditions in Context B, but simple exposure to Context A (Group Ao/B−). In the test, only Group A+/B− showed significantly more magazine entries shortly after the single food pellet in Context A, compared to Context B. Implications for theories of occasion setting and context modulation are considered.

Introduction

Theories differ on how context affects responses to a conditioned stimulus (CS). For example, associative theories (Mackintosh, 1975, Pearce and Hall, 1980, Rescorla and Wagner, 1972) suggest the context's own excitatory or inhibitory associations with the unconditioned stimulus (US) summate with the CS. However, modulatory theories (Bouton, 1993, Miller and Schachtman, 1985, Swartzentruber, 1995) suggest that context can modulate CS responding relatively independently of the context's own associative strength. Although these two classes of theories are not exhaustive, the dichotomy between them will be emphasized in this paper.

Context modulation resembles occasion setting, in which a feature CS signals that a target cue will (in feature-positive training) or will not (in feature-negative training) be subsequently reinforced. Recent evidence suggests that a feature CS controls target cue responding relatively independently of the feature's own associations with the US, especially when the feature and target are presented serially (Holland, 1992, Lamoureux et al., 1998, Morell and Holland, 1993, Rescorla, 1985; but see Holland, 1989). Several researchers have noted the similarity between context–CS, and feature–target, relations and have shown that context can modulate CS responding in a manner similar to that shown in occasion setting (Bouton and Swartzentruber, 1986, Hall and Mondragón, 1998, Swartzentruber, 1991).

Typically, occasion setting studies have examined the modulation of CS cues (such as tones or lights). Recently, however, we have examined the modulation of US cues and have uncovered some important differences between US and CS modulation (Goddard, 1999a, Goddard and Holland, 1996, Goddard and Holland, 1997). Further, certain theories in occasion-setting have been challenged by the sometimes striking differences between US and CS modulation, and it is now becoming increasingly clear that the defining characteristics of occasion-setting, established almost exclusively with CS cues, may be more task specific than originally thought (Skinner et al., 1998).

In a recent series of studies (Goddard and McDowell, 2001), we extended our previous research on US modulation by examining the ability of context to modulate US signal value. In Goddard and McDowell (2001), rats received training in which, in Context A, a single food pellet was delivered; this single food pellet was then followed shortly after by three additional pellets. Training in Context A alternated with training in Context B in which subjects also received a single food pellet; this single food pellet, however, was not followed by three additional pellets until after a delay. Magazine entries shortly after the single food pellet constituted the dependent variable of interest. Notably, session length, and the number of USs delivered in both contexts, was equated, and the context identity was counterbalanced. These procedural controls made it less likely that context modulation of US signal value was due to differences in context identity or context conditioning.

According to associative theories (Mackintosh, 1975, Pearce and Hall, 1980, Rescorla and Wagner, 1972), magazine (or food cup) entries after the single food pellet should be identical in the two contexts. Since context conditioning is equated, contextual associative strength should be equivalent prior to, and after, the delivery of the single food pellet. However, according to modulatory theories (Bouton, 1993, Miller and Schachtman, 1985, Swartzentruber, 1995), subjects should learn that the single food pellet is followed shortly after by three additional pellets in Context A, but not B, and evidence this learning by showing significantly more magazine entries shortly after the single food pellet in Context A, compared to Context B.

Results from Goddard and McDowell (2001) supported modulatory theories. Subjects showed significantly more magazine entries shortly after the single food pellet in Context A, compared to Context B, and extinguishing the associative strength of either context had little effect on context modulation. In addition to providing an important test of associative and modulatory theories, the study by Goddard and McDowell (2001) provided the first empirical demonstration that context can modulate US signal value.

However, in Goddard and McDowell (2001), context modulation of US signal value was established by explicit training. In Bouton and Swartzentruber (1986), context modulation of CS signal value was also established by explicit training in which Context A signaled that a CS was to be followed by a US, but Context B signaled that a CS was not to be followed by a US.

There is evidence, however, that context modulation may not require explicit training. For example, in latent inhibition, CS pre-exposure retards subsequent conditioning, but only if pre-exposure and conditioning are given in the same context. If pre-exposure is given in one context and conditioning in another, latent inhibition is attenuated or even abolished (Hall and Honey, 1989, Lovibond et al., 1984). Although no explicit training is given in these experiments, the context acquires control over what is learned in pre-exposure, since the effects of pre-exposure are evident only in the context in which CS pre-exposure was given. There are also examples of the failure of conditioned responding to transfer from one context to another, even when conditioning does not involve explicit training. For example, Hall and Honey (1989) showed that when rat subjects received CS A training in Context X and CS B training in Context Y, subjects showed greater conditioned responding when tested with previously presented (AX and BY), rather than different (AY and BX), context-stimulus pairings. Since no explicit training had been given (that is, subjects did not receive nonreinforced presentations of CS A in Context Y or CS B in Context X), context acquired conditional control in the absence of explicit training (see also Bonardi, 1992, Bonardi, 1998).

In Goddard and McDowell (2001), differential responding in Contexts A and B was established by explicit training. However, differential responding may have been established with nonexplicit training in which subjects received only the conditions that were in place in one context. The present experiment tested this alternative using three groups: Group A+/B−, Group A+/Bo, and Group Ao/B−. Subjects in Group A+/B− received explicit training that resembled the training used in Goddard and McDowell (2001). Subjects in Group A+/Bo received the identical conditions in Context A but simple exposure to Context B, and subjects in Group Ao/B− received the identical conditions in Context B but simple exposure to Context A.

After training, all subjects received four testing sessions in which a single food pellet was delivered in both contexts, and magazine entries were recorded. If, in Goddard and McDowell (2001), context modulation of US signal value requires explicit training, only those subjects in Group A+/B− should show differential responding in the two contexts. However, if context modulation of US signal value does not require explicit training, subjects receiving only the conditions in place in Context A (Group A+/Bo) or Context B (Group Ao/B−) should show differential responding in the two contexts. Notably, the present experiment, like Goddard and McDowell (2001), equated session length and the number of USs delivered in both contexts, in addition to counterbalancing context identity.

Section snippets

Subjects

The subjects were 24 individually housed female Sprague–Dawley rats obtained from the Charles River Company (St. Constant, Quebec, Canada). The subjects were about 115 days old at the beginning of the experiment and had not been involved in previous research. The animal holding room was maintained on a 14/10 h light/dark cycle with the lights on at 6:00 a.m. and off at 8:00 p.m. Subjects were maintained at 80% of their ad-libitum weight, by measured feedings of Purina Rat Chow at the end of a

Results

Averaged over the four testing sessions, subjects in Groups A+/B−, A+/Bo, and Ao/B− showed 3.1, 1.8, and 1.9 magazine entries/min before the single pellet in Context A and 2.0, 2.0, and 2.5 magazine entries/min, respectively, before the single pellet in Context B. These differences were not statistically significant (all Ts>3). In addition, there were no reliable group differences in responding before the single pellet in Contexts A and B (all Us>13).

Responding after the single food pellet,

Discussion

The results from Group A+/B− replicate the results obtained in Goddard and McDowell (2001). Subjects showed significantly more magazine entries shortly after the single food pellet in Context A (in which three additional pellets would shortly follow) compared to Context B (in which the delivery of three additional pellets was delayed). Notably, the present experiment, like that of Goddard and McDowell (2001), carefully equated session length, and US number, in each of two counterbalanced

Acknowledgements

This research was supported by Grant OGP 42025 from the Natural Sciences and Engineering Research Council of Canada. I thank Ken Lerette for his technical assistance and Marilyn MacLeod for helping score behavior.

References (30)

  • M.E. Bouton et al.

    Analysis of the associative and occasion-setting properties of contexts participating in a Pavlovian discrimination

    Journal of Experimental Psychology: Animal Behavior Processes

    (1986)
  • M.J. Goddard

    Effect of US signal value on blocking of a CS–US association

    Journal of Experimental Psychology: Animal Behavior Processes

    (1996)
  • M.J. Goddard et al.

    Type of feature affects transfer in operant serial feature-positive discriminations

    Animal Learning and Behavior

    (1996)
  • M.J. Goddard et al.

    Context modulation of US signal value

    The Quarterly Journal of Experimental Psychology

    (2001)
  • G. Hall et al.

    Contextual effects in conditioning, latent inhibition, and habituation: associative and retrieval functions of contextual cues

    Journal of Experimental Psychology: Animal Behavior Processes

    (1989)
  • Cited by (3)

    • Adaptive and maladaptive psychobiological responses to severe psychological stress: Implications for the discovery of novel pharmacotherapy

      2004, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      For example, trained soldiers will respond differently to a loud sound on the battlefield or when home on leave. Thus, response to CS (as well as to US) is context sensitive, as demonstrated in animals by the use of context modulation and occasion setting paradigms [36]. The intense fear response of patients with PTSD to CS regardless of context suggests impairment in their capacity to integrate context and content stimuli, illustrating how fear conditioning completely loses its adaptive function.

    • Occasion Setting

      2019, Behavioral Neuroscience
    View full text