criticisms of the satiety hypothesis as an explanation for within-session decreases in responding

347

JOURNAL OF THE EXPERIMENTAL ANALYSIS OF BEHAVIOR 2000, 74, 347–361 NUMBER 3 (NOVEMBER)

CRITICISMS OF THE SATIETY HYPOTHESIS ASAN EXPLANATION FOR WITHIN-SESSION

DECREASES IN RESPONDING

FRANCES K. MCSWEENEY AND ERIC S. MURPHY

WASHINGTON STATE UNIVERSITY

The authors of four papers recently reported that satiation provides a better explanation than ha-bituation for within-session decreases in conditioned responding. Several arguments question thisconclusion. First, the contribution of habituation to within-session changes in responding seemsclearly established. Information that is consistent with habituation, but that is difficult to reconcilewith satiation, is not adequately addressed. Second, the limited evidence offered in support of sati-ation is ambiguous because the results are just as compatible with habituation as with other satietyvariables. Finally, the term satiation is used in an intuitive way that is sometimes contradicted byresearch about the termination of ingestion. Use of the technical term satiation in a way that differsfrom its conventional usage will only isolate operant psychology from other areas of psychologicalresearch.

Key words: within-session changes in response rates, satiation, habituation, arousal, sensitization

Rate of responding often increases andthen decreases within a session when subjectsrespond on conditioning procedures (e.g.,McSweeney, 1992). These within-sessionchanges in responding are produced primar-ily by changes in the effectiveness of the re-inforcer with its repeated presentation in thesession (e.g., McSweeney, Hinson, & Cannon,1996; McSweeney, Weatherly, & Swindell,1996a). As a result, two ideas have competedto explain the within-session decreases in re-sponding. Killeen (e.g., 1995) argued that an-imals become satiated to repeatedly present-ed reinforcers (the satiety hypothesis).McSweeney et al. (1996) argued that subjectshabituate to aspects of the experimental pro-cedure that are presented repeatedly (e.g., re-inforcers) or for a prolonged time (e.g., thecontext; the habituation hypothesis). Both sa-tiation and habituation are assumed to re-duce the ability of reinforcers to support con-ditioned responding. The satiety andhabituation hypotheses are not entirely in-dependent. As will be discussed below, habit-uation is thought to be one of many factors

Preparation of this article was partially supported byGrant RO1 MH61720 from the National Institute of Men-tal Health. We thank Kenjiro Aoyama, Benjamin Kowal,Cam L. Melville, John Roll, Samantha Swindell, and Jef-frey N. Weatherly for their comments on an earlier ver-sion of this manuscript.

Correspondence about this paper should be addressedto Frances K. McSweeney, Department of Psychology,Washington State University, Pullman, Washington99164-4820 (E-mail: [email protected]).

that contribute to satiety (e.g., Swithers &Hall, 1994). Nevertheless, distinguishing be-tween the contribution of habituation andthe contributions of other satiety factors maybe important for both theoretical and ap-plied reasons.

From a theoretical standpoint, within-ses-sion changes in operant responding repre-sent a large, reliable, and general behavioraleffect (e.g., McSweeney & Roll, 1993). There-fore, a complete theory of operant respond-ing must accurately characterize the factorsthat produce those changes. In addition,within-session changes may eventually con-tribute to explaining many poorly under-stood phenomena in the conditioning litera-ture. Among these phenomena are extinction(McSweeney et al., 1996; McSweeney, Swin-dell, & Weatherly, 1999), multiple-schedulebehavioral contrast (McSweeney & Weatherly,1998), decreases in response rates at highrates of reinforcement (McSweeney, 1992),the contribution of economic factors to per-formance (McSweeney, Swindell, & Weather-ly, 1996b), and the regulation of motivatedbehavior (McSweeney & Swindell, 1999b; Roll& McSweeney, 1999). However, the implica-tions of within-session changes for these phe-nomena will differ if the satiety rather thanthe habituation hypothesis is correct. Multi-ple-schedule behavioral contrast will be usedto illustrate this point.

Traditionally, behavioral contrast refers toan inverse relation between the rate of re-

348 FRANCES K. MCSWEENEY and ERIC S. MURPHY

sponding during one component of a multi-ple schedule and reinforcement in the othercomponent (e.g., McSweeney & Norman,1979). For example, a multiple variable-inter-val (VI) 60-s VI 60-s schedule might bechanged to a multiple VI 60-s extinctionschedule. If rate of responding during the VI60-s component increased with this worsen-ing of alternative reinforcement, the increasein response rate would be labeled positive con-trast.

McSweeney and Weatherly (1998) ex-plained positive contrast by arguing that re-inforcers retain their effectiveness betteracross the session (less habituation or satia-tion) when only approximately 30 reinforcersare presented per hour during the multipleVI 60-s extinction schedule than when ap-proximately 60 reinforcers are presented perhour during the multiple VI 60-s VI 60-sschedule. The more effective reinforcers dur-ing the multiple VI 60-s extinction schedulesupport a higher rate of responding, and fast-er responding is observed during its VI 60-scomponent than during the same compo-nent of the multiple VI 60-s VI 60-s schedule.

This explanation of contrast is simple andrelies only on phenomena that have beendemonstrated by independent research.However, the experimental tests of this ideawill differ if repeatedly presented reinforcerslose their effectiveness through satiation rath-er than through habituation. If the habitua-tion hypothesis is correct, then manipulationsthat slow habituation should produce positivecontrast. For example, positive contrastshould be produced by manipulations such asreducing exposure to the stimulus by decreas-ing its size (e.g., Thompson & Spencer,1966), by introducing dishabituating stimulisuch as lights or noises (e.g., Thompson &Spencer, 1966), or by making unpredictablechanges in the reinforcer (e.g., Swithers &Hall, 1994). On the other hand, if the satietyhypothesis is correct, then positive contrastshould be produced by manipulations that re-duce the contribution of other satiety vari-ables. Such manipulations include decreasingthe size or caloric content of the reinforcer(e.g., Mook, 1996).

From an applied standpoint, the loss of ef-fectiveness that occurs when reinforcers arerepeatedly presented may be undesirablewhen operant techniques are used to correct

behavioral problems. The manipulations thatcan be used to reduce this loss, or to restorereinforcer effectiveness after it has been lost,will differ depending on whether the habit-uation or satiety hypothesis is correct. Again,if habituation is responsible, then these ma-nipulations will include introducing dishabi-tuators. If other satiety variables are respon-sible, then these manipulations will includereducing reinforcer size or caloric content.

The authors of four papers recently arguedthat satiation provides a better explanationthan habituation for within-session decreasesin responding (Bizo, Bogdanov, & Killeen,1998; DeMarse, Killeen, & Baker, 1999; Hin-son & Tennison, 1999; Palya & Walter, 1997).The present paper examines their arguments.We begin by defining our terms, because adiscussion of definitions may help to clarifythe issues. We then argue that three problemsplague the satiety hypothesis. First, it is incon-sistent with many empirical results that areconsistent with the habituation hypothesis.Second, the evidence that has been offeredto support the satiety hypothesis is ambigu-ous, because it is just as compatible with ha-bituation as it is with other satiety variables.Third, the satiety hypothesis, as offered inthese papers, is based on an intuitive, not anempirical, understanding of satiety.

DEFINITIONS

The use of the terms satiation and habitua-tion in the operant literature should be con-sistent with past research on each of thesetopics. If our definitions are consistent, thenwe can use information gathered in the pastto identify factors that should alter within-ses-sion changes in operant responding. If ourdefinitions are not consistent, then our idio-syncratic language will isolate us from otherareas of psychological research. No definitionof either satiation or habituation would beuniversally accepted by all researchers onthese topics (e.g., Savory, 1988; Weingarten,1985). However, the definitions that followare compatible with the way these terms areused by the researchers who study them.

Habituation

Habituation is often defined as a decreasein responsiveness to a stimulus when thatstimulus is presented repeatedly or for a pro-

349CRITICISMS OF THE SATIETY HYPOTHESIS

Table 1

A tentative list of the empirical characteristics of habituation. Characteristics that are precededby asterisks were included in Thompson and Spencer’s (1966) list of these characteristics.Examples of references that establish the role of other characteristics are listed.

*1. Spontaneous recovery: Responsiveness to a habituated stimulus recovers when that stimulus is not presented fora time.

2. Stimulus specificity (e.g., Swithers & Hall, 1994; Whitlow, 1975): Habituation is disrupted by unpredictablechanges in the presented stimulus.

3. Variety effects (e.g., Broster & Rankin, 1994): Perhaps because of stimulus specificity, habituation occurs moreslowly to stimuli that are presented in a variable, rather than a fixed, manner (e.g., after variable, rather thanfixed, interstimulus intervals).

*4. Dishabituation: Presenting a strong, different, or extra stimulus restores responsiveness to a habituated stimulus.Although dishabituation is listed here as a characteristic of habituation, researchers disagree about whether thereturn of responsiveness occurs because habituation decreases (e.g., Marcus, Nolen, Rankin, & Carew, 1988) orbecause sensitization is added (e.g., Groves & Thompson, 1970; see the discussion of sensitization below).

*5. Dishabituation habituates: Repeated presentation of dishabituators reduces their ability to restore habituatedresponding.

*6. Stimulus rate: Faster rates of stimulus presentation yield faster and more pronounced habituation than slowerrates.

7. Stimulus rate and recovery: Spontaneous recovery may be faster after faster than after slower rates of stimuluspresentation (Staddon & Higa, 1996).

*8. Stimulus exposure: Responsiveness to a repeatedly presented stimulus decreases with increases in stimulus ex-posure.

9. Long-term habituation (e.g., Wagner, 1976): Spontaneous recovery may be incomplete. Some habituation islearned and persists over time.

*10. Repeated habituations: Perhaps because of long-term habituation, habituation may become more rapid withrepeated habituation followed by spontaneous recovery.

*11. Stimulus intensity: Habituation is sometimes, but not always (e.g., Groves & Thompson, 1970), faster and morepronounced for less intense than for more intense stimuli.

12. Generality (e.g., Thorpe, 1966): Habituation occurs for most, if not all, species of animals. It also occurs formost stimuli, including those that have no ingestive consequences (e.g., lights, noises). The exact rate of habit-uation differs depending on the species, the stimulus, the response used as a measure, and the individual subject(e.g., Hinde, 1970).

Habituation is often accompanied by ‘‘sensitization’’ (e.g., Groves & Thompson, 1970). Therefore, if habituationoccurs, then the following phenomena might also be observed:13. Sensitization by early-stimulus presentations (e.g., Groves & Thompson, 1970): An increase (sensitization), rather

than a decrease (habituation), in responsiveness may occur to a repeatedly presented stimulus during its firstfew presentations.

14. Sensitization by stimuli from another modality (e.g., Swithers & Hall, 1994): An increase in responsiveness to astimulus may be produced by the introduction of a stimulus from another modality (e.g., a light or noise). Bothsensitization and dishabituation (Characteristic 4) may involve the introduction of a stimulus from anothermodality. Results are conventionally described as dishabituation if the stimulus restores responsiveness to analready habituated stimulus and as sensitization if the stimulus from another modality increases respondingbefore substantial habituation occurs to the other stimulus (e.g., Marcus et al., 1988).

longed time (e.g., Groves & Thompson,1970). Thompson and Spencer (1966) alsoargued that conformity to a list of empiricalproperties could be used as an elaborate op-erational definition of habituation. Since thetime of their paper, some of the empiricalcharacteristics that they offered have beenquestioned (e.g., the effect of stimulus inten-sity, e.g., Groves & Thompson); some are con-sidered to be of less generality (e.g., habitu-ation below zero, e.g., Wyers, Peeke, & Herz,1973); and others have been added to the list(e.g, sensitization, Groves & Thompson; stim-ulus specificity, Swithers & Hall, 1994). Nev-

ertheless, in the habituation literature, a de-crease in responsiveness to a repeatedlypresented stimulus is usually considered to behabituation when it conforms to a generallyaccepted list of empirical properties (e.g.,Leaton & Tighe, 1976; Teyler, Chiana, Di-Scenna, & Roemer, 1984).

Table 1 contains a tentative list of the em-pirical characteristics of habituation. Again,although there is general agreement on theproperties of habituation (e.g., Baker & Tif-fany, 1985), individual researchers mightomit some of the characteristics listed in Ta-ble 1 or include others. Decreases in respon-


siveness are usually called habituation even ifthose decreases fail to show all of these char-acteristics. For example, most researcherswould consider the decline in gill withdrawalin Aplysia with the repeated presentation of atactile stimulus to be a clear example of ha-bituation. However, by one count, gill with-drawal shows only six of Thompson andSpencer’s (1966) original nine characteristicsof habituation (Kandel, Castellucci, Pinsker,& Kupfermann, 1970).

Satiation

Satiation usually refers to a decline in theconsumption of an ingestive stimulus (e.g.,food, water) with its repeated consumption.The factors that contribute to the decline inconsumption are called satiety factors. Lists ofempirically identified satiety factors appear inmany textbooks. For example, the index ofMook (1996) lists the topic of ‘‘satiety (satia-tion) of feeding’’ (p. 633). The pages re-ferred to in this citation (pp. 76–79) list thefollowing satiety factors for food: oral stimu-lation; distension of the stomach; distensionof the duodenum; increases in blood sugarlevel at the liver; and increases in cholecys-tokinin (CCK) in the blood. Again, agree-ment among all researchers on any particularlist of satiety factors would be imperfect, al-though it would be substantial. Mook’s text-book was used to provide the present exam-ple because his own research is devoted tounderstanding satiety.

The list of satiety factors would also dependon the ingestive stimulus that was studied(e.g., food vs. water). Throughout this paper,we will use satiety to food as an example be-cause most, but not all (e.g., McSweeney,Weatherly, & Swindell, 1995b), studies ofwithin-session changes in responding usefood as a reinforcer. However, the specificpredictions of the satiety hypothesis would besomewhat different if a different reinforcerwas used.

In the past, the study of satiety focusedmainly on postingestive factors. That is, it fo-cused on factors that occur after food hasbeen digested by the stomach. More recentresearch has recognized the importance oforal factors. Habituation to the sensory prop-erties of the ingestive stimulus is one of thoseoral factors (e.g., Epstein, Rodefer, Wisniew-ski, & Caggiula, 1992; McSweeney & Swindell,

1999b; Swithers & Hall, 1994). This may in-troduce confusion, because studies that allowan animal to consume food orally produceboth habituation to the food through expo-sure to its sensory properties and changes inother satiety factors as a result of ingestion.The contribution of these different factorsmust be separated before any changes in be-havior can be clearly attributed to a particularvariable.

Throughout this paper we assume thatthose who favor the satiety hypothesis arguethat satiety factors other than habituationplay a role in producing within-session chang-es in operant responding. If they assumedthat habituation was the primary contributorto within-session changes, then they wouldagree with the habituation hypothesis, butthey do not (e.g., Bizo et al., 1998; DeMarseet al., 1999; Hinson & Tennison, 1999; Palya& Walter, 1997).

Separating the Satiety andHabituation Hypotheses

Habituation is usually classified as an oralfactor because satiety research has shown thathabituation to oral stimuli contributes to thetermination of feeding in rats and humans(e.g., Epstein et al., 1992; Swithers & Hall,1994). The precise nature of the stimuli towhich subjects habituate when consumingfood, however, has not been experimentallyidentified. The emphasis on oral factors prob-ably arose because satiety research usuallyuses rats and humans as subjects. It seems rea-sonable that oral factors should contributefor species that rely on taste and smell to gov-ern ingestion rather than on other sensorysystems (e.g., vision). The emphasis on oralfactors may be less appropriate for other spe-cies. For example, highly visual animals, suchas birds, might habituate primarily to visualcharacteristics of food (e.g., its color or size).Even rats might habituate to stimuli that arenot oral. For example, although this is spec-ulative, the location of food might be part ofthe stimulus to which subjects habituate formany species.

On the assumption that subjects mainly ha-bituate to the oral properties of food whenthey consume that food, physiological manip-ulations may be used to separate the contri-bution of habituation from the contributionsof other satiety factors. For example, sham


feeding procedures that shunt food from thebody before it reaches the stomach shouldproduce habituation (e.g., to the taste andtexture of the food) without any contributionby postingestive satiety factors (e.g., stomachdistension, changes in CCK and blood glu-cose levels) because food never reaches thestomach. In contrast, intubation of food di-rectly into the stomach should produce post-ingestive satiety factors without any contribu-tion of oral factors because the food bypassesthe mouth. (See Mook, 1996, pp. 76–79, forstudies that used these methods to isolate sa-tiety factors.)

This physiological approach is not entirelysatisfactory, however. First, the assumptionthat subjects habituate only to the oral prop-erties of food when consuming that food maybe incorrect. Therefore, subjects may be ex-posed to, and habituate to, other sensoryproperties of food (e.g., its temperature)even when food bypasses the mouth. Second,several satiety factors might be located in themouth. In that case, studies showing that in-gestion terminated when food entered themouth but immediately exited the bodywould remain ambiguous. Ingestion mighthave been terminated by, for example, a fixednumber of bites rather than by habituation.Finally, the behavior of a physiologically al-tered animal may differ from that of the in-tact animal.

As a result, our research has focused pri-marily on a second means of separating thecontribution of habituation from that of oth-er satiety factors. The empirical properties ofhabituation can be used to identify situationsin which the predictions of habituation differfrom those of other satiety variables. For ex-ample, increasing the caloric content of thefood should increase blood glucose levels andtherefore, produce faster satiety for a calorie-regulating animal such as the rat (e.g.,Adolph, 1947; Hausmann, 1933). In contrast,increasing the caloric content of the foodmight reduce habituation if it increased stim-ulus intensity (i.e., perceived sweetness). Ha-bituation is sometimes (e.g., Thompson &Spencer, 1966), but not always (e.g., Groves& Thompson, 1970), slower for more thanfor less intense stimuli. Table 2 provides moreexamples of these situations.

EVIDENCE FAVORING THEHABITUATION OVER THE

SATIETY HY POTHESIS

McSweeney et al. (1996) and McSweeneyand Roll (1998) showed that the empiricalcharacteristics of within-session changes in re-sponding are strikingly similar to the empir-ical characteristics of habituation. Their ar-guments will not be repeated here, except tostate that all of the empirical characteristicsof habituation listed in Table 1 have beenshown for within-session changes in respond-ing except Characteristics 5 (dishabituationhabituates), 7 (spontaneous recovery is fasterafter stimuli are presented at higher rates),and 14 (sensitization by stimuli from anothermodality). These characteristics have yet tobe studied. Some demonstrations of the char-acteristics of habituation were true predic-tions because they were confirmed only afterthe habituation hypothesis predicted them.Researchers have applied the term habituationto phenomena that share as few as three ofthe empirical characteristics of habituation(e.g., Eisenstein & Peretz, 1973). Becauseconformity to the empirical characteristics ofhabituation is used as an operational defini-tion of the term (e.g., Thompson & Spencer,1966), the contribution of habituation towithin-session changes in responding hasbeen established by typical standards of re-searchers who study habituation.

Do other satiety variables also contribute toproducing the within-session changes in re-sponding? Table 2 summarizes some evidencethat bears on this question. It presents a find-ing, lists some of the studies that support thatfinding, and then classifies whether the find-ing is consistent with the habituation hypoth-esis, with the satiety hypothesis, or with both.Asterisks indicate findings on which the in-terpretation listed in Table 2 disagrees withthe interpretation given in one of the fourpapers that favor the satiety hypothesis (Bizoet al., 1998; DeMarse et al., 1999; Hinson &Tennison, 1999; Palya & Walter, 1997). Table2 does not provide a complete summary ofresearch on within-session changes in re-sponding. Many such studies do not bear onthe distinction because they preceded the ar-guments over satiation versus habituation. Ta-ble 2 presents only those studies that may be


Table 2

Findings that have been reported for within-session changes in responding; some of the ref-erences that support that conclusion; and whether the conclusion is predicted by the habit-uation or satiety hypotheses. Asterisks indicate findings for which our classification of theevidence differs from that given by the authors of at least one study that reports that finding.When these studies are consistent with a characteristic of habituation as listed in Table 1, thenumber of that characteristic is given in parentheses after the finding.

FindingHabitua-

tion Satiation

Retrospective, not prospective, factors produce the decreases in responding x xMcSweeney, Weatherly, and Swindell (1995a)

A decrease in reinforcer effectiveness contributes to the decreases in responding x xMcSweeney, Weatherly, and Swindell (1996a)

The decreases are usually steeper for higher than for lower rates of reinforcement (6) x xMcSweeney (1992)McSweeney, Roll, and Cannon (1994)McSweeney, Roll, and Weatherly (1994)McSweeney and Swindell (1999a)McSweeney, Swindell, and Weatherly (1996a)McSweeney, Weatherly, and Swindell (1995b)McSweeney, Weatherly, and Swindell (1996b)

*The decreases are steeper for smaller or smaller cropped animals x xBizo, Bogdanov, and Killeen (1998)DeMarse, Killeen, and Baker (1999)Hinson and Tennison (1999)Palya and Walter (1997)

*The decreases may be steeper for larger (longer) than for smaller (shorter) food (8) x xBizo et al. (1998)Cannon and McSweeney (1995)DeMarse et al. (1999)Palya and Walter (1997)Roll, McSweeney, Johnson and Weatherly (1995)

*The decreases may differ for qualitatively different reinforcers (12) x xBizo et al. (1998)McSweeney, Swindell, and Weatherly (1996b)McSweeney, Weatherly, and Swindell (1996a)

*The decreases are steeper when animals are prefed and the food is eaten (8) x xDeMarse et al. (1999)

The decreases are not changed by prefeedings when food is intubated (8) xRoll et al. (1995)

*Frequent and unpredictable changes in external stimuli reduce the decreases (4) xHinson and Tennison (1999)

The decreases may be steeper for lower calorie (or less) than for higher calorie (or more)food (11) xMelville, Rue, Rybiski, and Weatherly (1997)Melville and Weatherly (1996)Weatherly, McSweeney, and Swindell (1997)

Increasing the caloric content of food may not steepen the decrease xRoll et al.

An increase in the amount of delivered food may increase, rather than decrease, responserate for the original reinforcer (4) xAoyama and McSweeney (in press)Ernst (2000)McSweeney and Roll (1998)


Table 2

(Continued)

FindingHabitua-

tion Satiation

Changing the reinforcer increases response rate even if the amount or value of reinforce-ment increases with the change (2) xErnst (2000)Rybiski (1996)

Increasing reinforcer variety reduces the steepness of the decreasea (3) xAoyama and McSweeney (in press)Ernst (2000)Facon and Darge (1996)McSweeney, Hinson, and Cannon (1996)Melville et al. (1997)

The decreases may occur for noningestive reinforcers (e.g., lights) (12) xFacon and Darge (1996)Kish (1966) for a review

Repeatedly delivered aversive stimuli also lose their effectiveness (12) xAzrin (1960)Jerome, Moody, Connor, and Ryan (1958)

The decreases may occur when no reinforcers are presented (12) xMcSweeney, Swindell, and Weatherly (1999)Schoenfeld, Antonitis, and Bersh (1950)

The decreases may change when the total food delivered is held constant xWeatherly, McSweeney, and Swindell (1995)

Changing food deprivation may not change the late-session decrease in responding for food xRoll et al. (1995)a Introducing variety slows both habituation and satiation. For example, habituation may be slower when the stimuli

are presented at variable, rather than at fixed, interstimulus intervals (Broseter & Rankin, 1994). Animals also eat ordrink more when given access to a variety of foods or liquids than when given access to only one (Berry, Beatty, &Klesges, 1985; Clifton, Burton, & Sharp, 1987; Rolls, Rolls, & Rowe, 1980; Rolls et al., 1981; Treit, Spetch, & Deutsch,1983). In spite of the effect of variety on satiation, we have not listed variety effects as a prediction of the satietyhypothesis because the variety effects in the satiety literature are produced by sensory factors such as habituationrather than by other satiety variables. For example, variety effects can occur when stimuli differ only in their sensoryproperties such as color or shape (Rolls, Rowe, & Rolls, 1982). They also occur when the food items have no nutritivevalue (Rolls, Wood, & Rolls, 1980).

useful in comparing the satiety and habitua-tion hypotheses.

Table 2 shows that many findings are com-patible with both the habituation and satietyhypotheses. Table 2 also lists 12 findings thatpit the predictions of habituation againstthose of other satiety variables. These predic-tions seem to fall into three general catego-ries. First, only the habituation hypothesiscorrectly predicts that within-session changesin responding may be observed when non-ingestive stimuli (e.g., lights, aversive stimuli)are repeatedly presented. Second, only thehabituation hypothesis correctly predicts thatresponse rates sometimes increase, ratherthan decrease, when the size or caloric con-

tent of the reinforcer increases (e.g., as a re-sult of dishabituation). Third, only the habit-uation hypothesis correctly predicts that thedecreases in operant responding can be al-tered by changing variables that should notalter other satiety variables (e.g., introducingstimuli from another modality).

Some of the findings reported in Table 2might be reconciled with the satiety hypoth-esis. For example, adding the concept ofarousal might allow the satiety hypothesis toaccount for the effect of manipulations suchas changing the nature of the reinforcer orintroducing frequent and unpredictablechanges in stimuli from another modality. Kil-leen and his colleagues have argued that the


repeated presentation of reinforcers may in-crease response rate by producing arousal(e.g., Killeen, Hanson, & Osborne, 1978) aswell as decrease response rate through satia-tion (e.g., Killeen, 1995). Just as habituationhas a companion process, sensitization, satia-tion may have a companion process, arousal.

However, sensitization has several advantag-es over arousal as an explanatory variable.First, researchers studying habituation werecompelled to postulate the concept of sensi-tization by the results of their studies (e.g.,Groves & Thompson, 1970). The postulationof arousal as a companion process for satia-tion arose from the study of operant respond-ing (e.g., Killeen et al., 1978), not the studyof satiety. Second, sensitization has been stud-ied extensively. Results of this research wereused to predict the manipulations that shouldreduce habituation and, therefore, alter with-in-session changes in responding. In contrast,the properties of arousal, as a companionprocess to satiety, have been studied muchless. Explicitly specifying the properties ofthis sort of arousal is important because theterm arousal has been used in a variety of con-flicting ways (e.g., Anderson, 1990; Duffy,1962; Neiss, 1988). Third, even if the prop-erties of arousal were established, those prop-erties would have to be similar to the prop-erties of sensitization to account for results inTable 2. If the two concepts make the samepredictions, then they may just name thesame process.

Many of the findings listed in Table 2 areincompatible with the satiety hypothesis be-cause they are inconsistent with research onthe termination of ingestion. These findingsmay be considered crucial experiments thatseparate the satiety and habituation hypoth-eses. Because many of the findings in Table2 were discussed by McSweeney and Roll(1998), we will describe only two here.

First, Melville, Rue, Rybiski, and Weatherly(1997) reported that rats’ responding some-times decreased more steeply late in the ses-sion when less concentrated, rather thanmore concentrated, sucrose solutions servedas reinforcers. Such a finding is consistentwith habituation, which may be faster andmore pronounced for less than for more in-tense stimuli (Characteristic 11, Table 1; e.g.,Thompson & Spencer, 1966). Decreasing thecaloric content of a sucrose solution might

decrease the intensity of its sweet taste. Thisfinding is the opposite of the prediction ofthe satiety hypothesis. Food ingestion shouldterminate faster for higher than for lower cal-orie solutions for animals, such as rats, thatregulate calories (e.g., Adolph, 1947; Haus-mann, 1933).

Although Bizo et al. (1998) questioned theinterpretation of data that are reported interms of proportions of total-session respons-es, Melville et al.’s (1997) conclusion survivesa conversion to response rates. For example,if the steepness of the decrease in operantresponse rate is measured by the differencebetween the number of responses emittedduring the 5-min intervals of the session thatcontained the highest and the lowest numberof responses, the decrease was 76.5 responsesfor the 1% sucrose solution, 66.0 responsesfor the 5% solution, and 59.0 responses forthe 15% sucrose solution. (See Melville &Weatherly, 1996, and Weatherly, McSweeney,& Swindell, 1997, for additional evidence thatthe decreases in operant responding may besteeper when less rather than more food isgiven.)

Second, McSweeney and Roll (1998) re-ported that changing the reinforcer for abrief time late in the session increased re-sponse rate once the orginal reinforcer wasrestored. The increase in responding oc-curred regardless of whether the change wasan increase or a decrease in the amount ofreinforcement delivered and regardless ofwhether the change produced an increase ora decrease in response rate while it was ineffect. These results are consistent with ha-bituation. Changes in the reinforcer shouldproduce dishabituation (Characteristic 4, Ta-ble 1; e.g., Thompson & Spencer, 1966). Be-cause dishabituation should restore respon-siveness to the habituated stimulus, responserate should increase. This finding contradictswhat would be expected from satiety factorssuch as stomach distension and blood glucoselevels. Providing more reinforcers should in-crease both of these factors and decrease, notincrease, response rate. These results werereplicated by Aoyama and McSweeney (inpress) and by Ernst (2000) using fixed- andvariable-ratio schedules as the baseline sched-ules of reinforcement. Therefore, the resultshave some generality and are not explainedby the disinhibition of any inhibition that


might occur during the fixed-interval sched-ule baselines used by McSweeney and Roll(e.g., Staddon, 1967).

Some other results that favor the habitua-tion over the satiety hypothesis are equallydifficult to dismiss. Table 2 shows that severalof these conclusions are supported by morethan one study. The studies often come fromdifferent laboratories. The studies also oftenexamine the behavior of both rats and pi-geons responding in several different condi-tions. Examining the behavior of two specieseliminates many potential procedural criti-cisms. For example, Bizo et al. (1998) ques-tioned whether our food hoppers for pigeonsreally provided different access to food whenreinforcer duration was varied. Althoughtheir criticism might explain why we some-times failed to find a change in the within-session response pattern when we manipulat-ed reinforcer size for pigeons (e.g., Cannon& McSweeney, 1995), it cannot explain whywe found similar results for rats (e.g., Roll,McSweeney, Johnson, & Weatherly, 1995).

ARGUMENTS FOR THESATIETY HY POTHESIS ARE

COMPATIBLE WITHHABITUATION

The findings in Table 2 that are precededby an asterisk have been interpreted as sup-porting the satiety hypothesis by Bizo et al.(1998), DeMarse et al. (1999), Palya and Wal-ter (1997), or Hinson and Tennison (1999).However, these results are at least as consis-tent with the habituation hypothesis as theyare with the satiety hypothesis.

Crop Capacity Correlated withthe Decrease in Responding

Three studies measured the amount thatsubjects ate under free-feeding conditions(Bizo et al., 1998; DeMarse et al., 1999) orclose to free-feeding conditions (DeMarse etal.; Hinson & Tennison, 1999). The amountconsumed was labeled capacity or crop capacity.It was assumed that the same amount of foodshould produce more satiety in birds thathave a smaller capacity. If satiation producesthe decreases in operant response rates, thosedecreases should be steeper for subjects withsmaller capacities. This prediction was con-firmed.

In a related argument, Palya and Walter(1997) showed that within-session decreasesin responding are steeper for smaller than forlarger birds. They argued that this result isconsistent with the satiety hypothesis if small-er birds became satiated more quickly thanlarger birds. Palya and Walter’s finding willnot be directly addressed because Bizo et al.(1998) and DeMarse et al. (1999) failed toreplicate it. However, the following criticismapplies to Palya and Walter’s reasoning as wellas to that of the other studies.

The reasoning behind these arguments isflawed. To begin with, the authors interpreta correlation (i.e., that body size or amounteaten under free-feeding conditions is cor-related with the steepness of the decrease inoperant response rate) as causation (e.g.,‘‘Satiation Causes Within-Session Decreasesin Instrumental Responding’’ is the title ofthe article by Bizo et al., 1998). Drawing caus-al conclusions from correlational data is log-ically incorrect. When two variables are cor-related, either one could cause the other orthe correlation could be caused by a thirdvariable.

It has been argued that statistical relations,including correlations, may suggest causationif certain conditions are met (e.g., Susser,1973). Although it is beyond the scope of thepresent paper, these conditions are strict(e.g., Susser, p. 142) and do not appear to bemet by the data presented for the satiety hy-pothesis. Nevertheless, assume that these cor-relations can be interpreted as causation.Even in that case, the causal variable is just aslikely to be habituation as it is to be othersatiety variables. When the authors argue thatcrop capacity is correlated with the decreasesin operant responding, they seem to be re-porting a relation between two independentphenomena (i.e., the physical size of the cropand the behavior of the animal). What theyhave actually done is measured feeding in twodifferent situations (the home cage and theoperant enclosure). Because the subjects con-sumed food orally when both crop capacityand the decreases in operant respondingwere measured, many factors were affected,including habituation because the subjectswere repeatedly exposed to the sensory prop-erties of food (e.g., Epstein et al., 1992; Swith-ers & Hall, 1994). Without other evidence,there is no logical basis for attributing the


correlation to any particular satiety factor.Habituation to the sensory properties of thefood provides as plausible an explanation ofthe correlation as any other satiety factor.

Steeper Decreases for Longer Reinforcers

Three studies showed that within-sessiondecreases in operant responding are steeperor operant response rates are slower whensubjects are given longer rather than shorterdurations of access to food per reinforcer(Bizo et al., 1998, Experiment 1; DeMarse etal., 1999, Experiment 2; Palya & Walter, 1997,Experiment 2). Again, however, as acknowl-edged by Bizo et al., habituation provides justas plausible an explanation as other satietyvariables. Subjects received greater exposureto the sensory properties of the food duringlonger than during shorter reinforcers. Ha-bituation usually increases with increases instimulus exposure (Characteristic 8, Table 1;e.g., Thompson & Spencer, 1966).

Different Grains Produce DifferentDecreases in Responding

Bizo et al. (1998, Experiment 2) reasonedthat if satiation produces decreases in oper-ant response rate, then those decreasesshould be steeper for larger than for smallergrains. They showed that decreases are steep-er for popcorn than for milo. Again, however,finding different decreases for differentgrains may be compatible with the habitua-tion hypothesis. The rate of habituation de-pends on the nature of the stimulus delivered(Characteristic 12, Table 1; e.g., Hinde,1970). Therefore, rate of habituation is likelyto differ for popcorn and milo.

Rate of habituation may differ for differentgrains, but one could still argue that the ha-bituation hypothesis does not explicitly pre-dict faster decreases for popcorn than formilo. This is correct, but Bizo et al. (1998)provide no evidence that the satiety hypoth-esis makes this prediction either. The studyby Ploog and Zeigler (1996) that they cite in-dicates that pigeons peck faster during con-ditioning procedures for larger than forsmaller grain reinforcers. It does not showthat pigeons terminate consumption (be-come satiated) faster for larger than for small-er grains or that this faster termination re-sults from a variable other than habituation(e.g., from crop or stomach distension).

Prefeeding Animals ReducesResponse Rate

DeMarse et al. (1999, Experiment 3) rea-soned that feeding subjects before the sessionshould increase satiety. They fed the pigeons0, 5, 15, or 25 g of milo prior to the sessionand reported that response rates decreasedwith increases in the amount prefed.

Again, these data are just as compatiblewith the habituation hypothesis as they arewith the satiety hypothesis. Because the foodconsumed before the session was similar tothe food used as a reinforcer, prefeeding ex-posed the subjects to the sensory propertiesof that food. Therefore, subjects should haveentered the session more habituated to thefood when they were prefed than when theywere not. Larger prefeedings should meanmore sensory exposure and, therefore, morehabituation (Characteristic 8, Table 1; e.g.,Thompson & Spencer, 1966).

Although Experiment 2 of Roll et al.(1995) was not discussed by DeMarse et al.(1999), this experiment also examined the ef-fect of prefeeding. The rats in Roll et al. re-sponded on a multiple VI 60-s VI 60-s sched-ule during baseline when no prefeedingswere given. The amount of condensed milkreinforcer consumed up to the point in thesession at which the peak response rate oc-curred was calculated for each rat to deter-mine the amount of food required to initiate‘‘satiation.’’ Subjects were then divided intotwo groups that received either 1.0 or 2.0times their predetermined amount of milkimmediately before the session. Unlike theDeMarse et al. study, the prefeeding was givenby intubating food directly into the stomach.Roll et al. reported that prefeeding the ratsthis way did not alter the within-session pat-tern of responding for either group.

Roll et al.’s (1995) results are consistentwith the habituation hypothesis, but they maynot be consistent with the satiety hypothesis.Because the prefeeding food was intubatedrather than eaten, the subjects could not ha-bituate to the oral properties of that food(e.g., its taste and texture). In contrast, intu-bations should produce some satiety becausefood placed in the stomach alters known sa-tiety factors such as stomach distension,blood glucose levels, and so on. Therefore,the prefeedings should steepen the within-


session decrease in responding if those de-creases are produced by satiety. A steepeningof the decreases was not observed.

Roll et al.’s (1995) results may be criticized.For example, intubation is an intrusive pro-cedure that may not correctly represent ananimal’s normal feeding. Nevertheless, theseresults deserve discussion because physiolog-ical interventions such as intubation are fre-quently used to separate the contributions ofpre- and postingestion satiety factors (e.g.,Mook, 1996). The habituation hypothesis isconsistent with the results of both DeMarseet al. (1999) and Roll et al. The satiety hy-pothesis is consistent only with the results ofDeMarse et al.

Hinson and Tennison (1999)

In a series of four experiments, Hinson andTennison (1999) failed to find within-sessiondecreases in responding when they providedrates of reinforcement similar to those that usu-ally produce these decreases (e.g., McSweeney,1992). They concluded that their results sup-port the satiety hypothesis because a measureof capacity (i.e., consumption) predicted thepresence of a decrease in responding whensuch a decrease did occur. As argued earlier,finding a correlation between consumption(i.e., capacity) and the decrease in respondingprovides no support for the satiety over the ha-bituation hypothesis. Once this argument is dis-missed, Hinson and Tennison’s results aremore consistent with the habituation hypothe-sis than with the satiety hypothesis.

Habituation provides an obvious explana-tion for Hinson and Tennison’s (1999) find-ings. They used a complicated discriminationprocedure in which visual stimuli changedapproximately every 5 to 30 s in differentconditions. The newly introduced stimuluswas unpredictable and was sometimes drawnfrom a large set (e.g., 10 or 12 stimuli). Inaddition, the stimuli were sometimes flicker-ing rather than constant lights. The introduc-tion of these visual stimuli should producedishabituation (Characteristic 4, Table 1) fora highly visual animal such as a pigeon. Al-though subjects may habituate to dishabitua-tors (Characteristic 5, Table 1), these partic-ular visual stimuli should retain theireffectiveness as dishabituators throughout thesession because stimulus changes were fre-quent and unpredictable (Characteristics 2

and 3, Table 1). Dishabituation by the fre-quently changing visual stimuli should reducehabituation to the reinforcers and reduce thesize of the within-session decrease in respond-ing. This was the result reported by Hinsonand Tennison.

In contrast, the satiety hypothesis provideslittle reason to expect different results in Hin-son and Tennison’s (1999) experiments thanin past studies that provided similar condi-tions of reinforcement. When the character-istics of the reinforcers (e.g., their size, rateof delivery, etc.) are held constant, satiety fac-tors such as stomach distension, blood glu-cose levels, and so on should also be constant.Introducing external stimuli, such as lights,should not alter these satiety variables.

Hinson and Tennison (1999) recognizedthat a factor such as dishabituation might ac-count for their results. They state that theirmultiple and rapidly changing stimuli mightmaintain ‘‘attention’’ to the stimulus andtherefore, eliminate habituation. However,their use of the term attention is questionable.The results of past research can be used topredict that dishabituation will occur in thissituation. Dishabituation is usually thought ofas either a release from habituation or theintroduction of sensitization (Table 1). It isnot equated with the maintenance of atten-tion, a concept with very different implica-tions. For example, attention is often as-sumed to be a limited resource, so thatgreater attention to some stimuli means lessattention to other stimuli (e.g., Hinson &Cannon, 1999). Dishabituation and sensitiza-tion are not limited in this way.

The Use of Proportions

Bizo et al. (1998) reasoned that the argu-ments supporting the habituation hypothesisare flawed because they are often based onproportions of total-session responses ratherthan on response rates. They are correct thatproportions may sometimes lead to differentconclusions than do response rates. Examplesmay be found in the Appendix of McSweeneyand Roll (1998). However, there is no a priorireason to assume that response rates providea ‘‘truer’’ description of behavior than do re-sponse proportions. Neither measure has log-ical priority because both are calculated bydividing number of responses by anothermeasure (session time or total-session re-


sponses). Instead, McSweeney and Roll sug-gest deciding between these measures by ask-ing which is the most orderly. They thenshowed that proportions, not response rates,bring order to the data.

In addition, the statistical tests reported inour studies were almost always applied to re-sponse rates rather than to proportions. Pro-portions are bounded and must sum to 1 inviolation of the assumptions of some statisti-cal tests. Our conclusions were based on theresults of statistical tests on response rates aswell as on a visual analysis of response pro-portions. Therefore, our conclusions almostalways apply to data analyzed as responserates, not just as response proportions.

AN INTUITIVE TREATMENTOF SATIATION

Operant researchers should use the termssatiation and habituation in a manner that isconsistent with the research from which theterms arise. Instead, the arguments in sup-port of the satiety hypothesis seem to bebased on intuition rather than on an empir-ical understanding of satiety.

To begin with, the experimental manipu-lations used differ from those usually used toidentify satiety factors. For example, the dif-ference in the within-session decreases in re-sponding for popcorn and milo (Bizo et al.,1998) might arise from any of the many di-mensions on which popocorn and milo differ(weight, size, shape, color, texture, taste, ca-loric content, nutrient composition, topog-raphy of the response required for eating,etc.). Even if we did know what variable wasresponsible (e.g., size), we would not knowwhether that variable had its effect throughhabituation (e.g., subjects habituate faster tolarger grains because they provide greatersensory exposure) or through other satietyfactors (e.g., larger grains produce greatercrop or stomach distension). Better tests ofthe satiety hypothesis would directly manip-ulate variables that are clearly identified withparticular satiety factors (e.g., caloric content,stomach distension).

Research on the termination of food con-sumption also contradicts an argument of-fered by Bizo et al. (1998). Roll et al. (1995)showed that deprivation for food does not al-ter the within-session pattern of responding

for food reinforcers. They argued that such aresult questions satiation because deprivationshould alter the rate of satiety. Bizo et al. dis-agree. They state that ‘‘the slope of within-session responding should not be expected tochange when deprivation level is manipulat-ed. All deprivation does is change the initiallevel of responding’’ (p. 446). Because theydo not cite references, it is unclear what dataled to that conclusion. Contrary to their ar-gument, however, changing deprivation doesalter the changes in eating rate that occurwithin a meal (i.e., the slope), not just theinitial rate of eating (e.g., Bousfield, 1935;Bousfield & Elliott, 1934; Kissileff & Thorn-ton, 1982; Savory, 1988). In fact, at least onestudy reported that deprivation alters theslope but not the initial rate of eating, whichis the opposite of Bizo et al.’s conclusion(e.g., Aoyama, 2000).

As a final point, the factors that producesatiety differ for different species. For exam-ple, calories play no role in regulating the con-sumption of food by the blowfly (e.g., Dethier,1962, 1976) even though they play a large rolein regulating consumption by rats and hu-mans (e.g., Adolph, 1947; Hausmann, 1933).It is therefore noteworthy that all four of thestudies that support the satiety hypothesis usedpigeons as subjects. On the surface, pigeonsseem to be a natural choice because they havebeen used frequently in studies of within-ses-sion changes in responding. But pigeons arenot a good choice of subject if one is primarilyinterested in the role of satiety in producingthose changes. Research on the termination ofingestion usually examines the behavior of ratsand humans. Less is known about the termi-nation of ingestion in the pigeon. The partic-ular factors that contribute to satiety must beindentified before it can be asked if those fac-tors also contribute to the production of with-in-session changes in responding. As a result,pigeons are an odd choice of subject if inves-tigators are interested in showing that satietyfactors contribute to within-session decreasesin operant responding.

PROBLEMS WITHSENSITIZATION-HABITUATION

Although sensitization-habituation has ad-vantages over arousal-satiation as an expla-nation for the within-session changes in op-


erant responding, it also has some flaws. Tobegin with, the hypothesis has mechanismsthat both increase (e.g., sensitization) and de-crease (habituation) response rate. Postulat-ing two factors gives a theory a lot of latitude.However, postulating two factors may be nec-essary to explain the frequently bitonic with-in-session changes in responding. In addi-tion, research on sensitization-habituationconstrains the conditions under which re-sponse rates may increase. Increases may oc-cur when only a few stimuli have been pre-sented (sensitization); after the presentationof a strong, different, or extra stimulus (dis-habituation); after a change in the nature ofthe stimulus (violations of stimulus specifici-ty); or when the stimulus has not been pre-sented for a time (spontaneous recovery). Ifincreases occur under other conditions or failto occur under these conditions, then thetheory is wrong or at least incomplete.

The sensitization-habituation model is alsoan empirical generalization rather than a for-mal theory. It states only that whatever pro-duces sensitization and habituation to stimuli,such as lights and tones, also produces thedynamic changes in reinforcer effectivenessthat occur within a session. It leaves thoseprocesses unidentified. An empirical ap-proach is necessary at present because theleading models of habituation (e.g., Sokolov,1963; Wagner, 1976) have been severely crit-icized (e.g., Mackintosh, 1987; Staddon &Higa, 1996). An empirical approach also hasan important advantage: Theory-based re-search becomes less useful when the theoryon which it is based is questioned. Neverthe-less, our hypothesis would have more contentif a generally accepted theory of habituationwere available.

Many questions have not yet been an-swered about sensitization and habituation.For example, rate of operant responding maydecrease within the session for one subject,but response rate first increases and then de-creases for another subject. Similar individualdifferences in rates of sensitization and ha-bituation are frequently reported (e.g.,Hinde, 1970). However, the factors that de-termine why sensitization or habituation isstronger for one subject than for another re-main to be identified.

Finally, the sensitization-habituation expla-nation may be incomplete. For example, sa-

tiety factors other than habituation may even-tually be shown to play a role in regulatingthe within-session changes in conditioned re-sponding. At present, however, the only vari-ables that have been shown to play a role re-main sensitization and habituation. Table 2shows that the role of these variables is wellsupported. In contrast, the arguments forother satiety variables are limited and ambig-uous. In addition, some of the evidence pre-sented in Table 2 suggests that the contribu-tion of habituation is large enough toovercome the contribution of other satietyfactors. Therefore, other satiety variables mayplay a smaller role in producing within-ses-sion changes in responding than intuitionmight initially suggest.

REFERENCES

Adolph, E. F. (1947). Urges to eat and drink in rats.American Journal of Physiology, 151, 110–125.

Anderson, K. J. (1990). Arousal and the inverted-u hy-pothesis: A critique Neiss’s ‘‘Reconceptualizing Arous-al.’’ Psychological Bulletin, 107, 96–100.

Aoyama, K. (2000). Effects of hunger state on within-session response decreases under CRF schedule.Learning and Motivation, 31, 1–20.

Aoyama, K., & McSweeney, F. K. (in press). Habituationmay contribute to within-session decreases underhigh-rate schedules of reinforcement. Animal Learning& Behavior.

Azrin, N. H. (1960). Sequential effects of punishment.Science, 131, 605–606.

Baker, T. B., & Tiffany, S. T. (1985). Morphine toleranceas habituation. Psychological Review, 92, 78–108.

Berry, S. L., Beatty, W. W., & Klesges, R. C. (1985). Sen-sory and social influences on ice cream consumptionby males and females in a laboratory setting. Appetite,6, 41–45.

Bizo, L. A., Bogdanov, S. V., & Killeen, P. R. (1998). Sa-tiation causes within-session decreases in instrumentalresponding. Journal of Experimental Psychology: AnimalBehavior Processes, 24, 439–452.

Bousfield, W. A. (1935). Quantitative indices of the ef-fects of fasting on eating-behavior. Journal of GeneticPsychology, 46, 476–479.

Bousfield, W. A., & Elliott, M. H. (1934). The effect fast-ing on the eating-behavior of rats. Journal of GeneticPsychology, 45, 227–237.

Broster, B. S., & Rankin, C. H. (1994). Effects of chang-ing interstimulus interval during habituation in Cae-norhabditis elegans. Behavioral Neuroscience, 108, 1019–1029.

Cannon, C. B., & McSweeney, F. K. (1995). Within-ses-sion changes in responding when rate and durationof reinforcement vary. Behavioural Processes, 34, 285–292.

Clifton, P. G., Burton, M. J., & Sharp, C. (1987). Rapidloss of stimulus-specific satiety after consumption of asecond food. Appetite, 9, 149–156.


DeMarse, T. B., Killeen, P. R., & Baker, D. (1999). Sati-ation, capacity, and within-session responding. Journalof the Experimental Analysis of Behavior, 72, 407–423.

Dethier, V. G. (1962). To know a fly. San Francisco: Hold-en-Day.

Dethier, V. G. (1976). The hungry fly: A physiological studyof the behavior associated with feeding. Cambridge, MA:Harvard University Press.

Duffy, E. (1962). Activation and behavior. New York: Wiley.Eisenstein, E. M., & Peretz, B. (1973). Comparative as-

pects of habituation in invertebrates. In H. V. S. Peeke& M. J. Herz (Eds.), Habituation: Vol. 2. Physiologicalsubstrates (pp. 1–31). New York: Academic Press.

Epstein, L. H., Rodefer, J. S., Wisniewski, L., & Caggiula,A. R. (1992). Habituation and dishabituation of hu-man salivary response. Physiology & Behavior, 51, 945–950.

Ernst, M. M. (2000). Within-session changes in operant re-sponding for food in humans. Unpublished doctoral dis-sertation, State University of New York, University atBuffalo.

Facon, B., & Darge, T. M. (1996). Evaluation of toy var-iation on engagement in a leisure activity of two chil-dren with profound multiple handicaps. PsychologicalReports, 79, 203–210.

Groves, P. M., & Thompson, R. F. (1970). Habituation:A dual-process theory. Psychological Review, 77, 419–450.

Hausmann, M. F. (1933). The behavior of albino rats inchoosing foods: II. Differentiation between sugar andsaccharin. Journal of Comparative Psychology, 15, 419–428.

Hinde, R. A. (1970). Behavioral habituation. In G. Horn& R. A. Hinde (Eds.), Short-term changes in neural ac-tivity and behavior (pp. 3–40). Cambridge, England:Cambridge University Press.

Hinson, J. M., & Cannon, C. B. (1999). Stimulus spacing,attention, and dimensional contrast. Animal Learning& Behavior, 27, 472–480.

Hinson, J. M., & Tennison, L. R. (1999). Within-sessionanalysis of visual discrimination. Journal of the Experi-mental Analysis of Behavior, 72, 385–405.

Jerome, E. A., Moody, J. A., Connor, T. J., & Ryan, J.(1958). Intensity of illumination and the rate of re-sponding in a multiple-door situation. Journal of Com-parative and Physiological Psychology, 51, 47–49.

Kandel, E., Castellucci, V., Pinsker, H., & Kupfermann, I.(1970). The role of synaptic plasticity in the short-term modification of behavior. In G. Horn & R. A.Hinde (Eds.), Short-term changes in neural activity andbehavior (pp. 281–322). Cambridge, England: Cam-bridge University Press.

Killeen, P. R. (1995). Economics, ecologics, and me-chanics: The dynamics of responding under condi-tions of varying motivation. Journal of the ExperimentalAnalysis of Behavior, 64, 405–431.

Killeen, P. R., Hanson, S. J., & Osborne, S. R. (1978).Arousal: Its genesis and manifestation as responserate. Psychological Review, 85, 571–581.

Kish, G. B. (1966). Studies of sensory reinforcement. InW. K. Honig (Ed.), Operant behavior: Areas of researchand application (pp. 109–159). New York: Appleton-Century-Crofts.

Kissileff, H. R., & Thornton, J. (1982). Facilitation andinhibition in the cumulative food intake curve inman. In A. R. Morrison & P. L. Strick (Eds.), Changing

concepts of the nervous system: Proceedings of the first insti-tute of neurological sciences symposium in neurobiology (pp.585–607). New York: Academic Press.

Leaton, R. N., & Tighe, T. J. (1976). Comparisons be-tween habituation research at the developmental andanimal-neurophysiological levels. In T. J. Tighe & R.N. Leaton (Eds.), Habituation: Perspectives from child de-velopment, animal behavior and neurophysiology (pp. 321–340). New York: Wiley.

Mackintosh, N. J. (1987). Neurobiology, psychology andhabituation. Behaviour Research and Therapy, 25, 81–97.

Marcus, E. A., Nolen, T. G., Rankin, C. H., & Carew, T.J. (1988, July 8). Behavioral dissociation of dishabit-uation, sensitization, and inhibition in Aplysia. Science,241, 210–213.

McSweeney, F. K. (1992). Rate of reinforcement and ses-sion duration as determinants of within-session pat-terns of responding. Animal Learning & Behavior, 20,160–169.

McSweeney, F. K., Hinson, J. M., & Cannon, C. B. (1996).Sensitization-habituation may occur during operantconditioning. Psychological Bulletin, 120, 256–271.

McSweeney, F. K., & Norman, W. D. (1979). Definingbehavioral contrast for multiple schedules. Journal ofthe Experimental Analysis of Behavior, 32, 457–461.

McSweeney, F. K., & Roll, J. M. (1993). Respondingchanges systematically within sessions during condi-tioning procedures. Journal of the Experimental Analysisof Behavior, 60, 621–640.

McSweeney, F. K., & Roll, J. M. (1998). Do animals sa-tiate or habituate to repeatedly presented reinforcers?Psychonomic Bulletin & Review, 5, 428–442.

McSweeney, F. K., Roll, J. M., & Cannon, C. B. (1994).The generality of within-session patterns of respond-ing: Rate of reinforcement and session length. AnimalLearning & Behavior, 22, 252–266.

McSweeney, F. K., Roll, J. M., & Weatherly, J. N. (1994).Within-session changes in responding during severalsimple schedules. Journal of the Experimental Analysis ofBehavior, 62, 109–132.

McSweeney, F. K., & Swindell, S. (1999a). Behavioraleconomics and within-session changes in responding.Journal of the Experimental Analysis of Behavior, 72, 355–371.

McSweeney, F. K., & Swindell, S. (1999b). General-pro-cess theories of motivation revisited: The role of ha-bituation. Psychological Bulletin, 125, 437–457.

McSweeney, F. K., Swindell, S., & Weatherly, J. N.(1996a). Within-session changes in responding dur-ing autoshaping and automaintenance procedures.Journal of the Experimental Analysis of Behavior, 66, 51–61.

McSweeney, F. K., Swindell, S., & Weatherly, J. N.(1996b). Within-session changes in responding dur-ing concurrent schedules with different reinforcers inthe components. Journal of the Experimental Analysis ofBehavior, 66, 369–390.

McSweeney, F. K., Swindell, S., & Weatherly, J. N. (1999).Within-session response patterns during variable in-terval, random reinforcement, and extinction proce-dures. Learning and Motivation, 30, 221–240.

McSweeney, F. K., & Weatherly, J. N. (1998). Habituationto the reinforcer may contribute to multiple-schedulebehavioral contrast. Journal of the Experimental Analysisof Behavior, 69, 199–221.

McSweeney, F. K., Weatherly, J. N., & Swindell, S.


(1995a). Prospective factors contribute little to with-in-session changes in responding. Psychonomic Bulletin& Review, 2, 234–238.

McSweeney, F. K., Weatherly, J. N., & Swindell, S.(1995b). Within-session changes in key and leverpressing for water during several multiple variable-in-terval schedules. Journal of the Experimental Analysis ofBehavior, 64, 75–94.

McSweeney, F. K., Weatherly, J. N., & Swindell, S.(1996a). Reinforcer value may change within exper-imental sessions. Psychonomic Bulletin & Review, 3, 372–375.

McSweeney, F. K., Weatherly, J. N., & Swindell, S.(1996b). Within-session changes in responding dur-ing concurrent variable-interval schedules. Journal ofthe Experimental Analysis of Behavior, 66, 75–95.

Melville, C. L., Rue, H. C., Rybiski, L. R., & Weatherly, J.N. (1997). Altering reinforcer variety or intensitychanges the within-session decrease in responding.Learning and Motivation, 28, 609–621.

Melville, C. L., & Weatherly, J. N. (1996). Within-sessionpatterns of responding when rats run in a T-maze.Behavioural Processes, 38, 89–102.

Mook, D. G. (1996). Motivation: The organization of action(2nd ed.). New York: W. W. Norton.

Neiss, R. (1988). Reconceptualizing arousal: Psychobio-logical states in motor performance. Psychological Bul-letin, 103, 345–366.

Palya, W. L., & Walter, D. E. (1997). Rate of a main-tained operant as a function of temporal positionwithin a session. Animal Learning & Behavior, 25, 291–300.

Ploog, B. O., & Zeigler, H. P. (1996). Effects of food-pellet size on rate, latency, and topography of auto-shaped key pecks and gapes in pigeons. Journal of theExperimental Analysis of Behavior, 65, 21–35.

Roll, J. M., & McSweeney, F. K. (1999). Within-sessionchanges in response rate: Implications for behavioralpharmacology. The Psychological Record, 49, 15–32.

Roll, J. M., McSweeney, F. K., Johnson, K. S., & Weatherly,J. N. (1995). Satiety contributes little to within-ses-sion decreases in responding. Learning and Motivation,26, 323–341.

Rolls, B. J., Rolls, E. T., & Rowe, E. A. (1980). Specificsatiety and its influence on feeding. Appetite, 1, 85–86.

Rolls, B. J., Rowe, E. A., & Rolls, E. T. (1982). How sen-sory properties of foods affect human feeding behav-ior. Physiology & Behavior, 29, 409–417.

Rolls, B. J., Rowe, E. A., Rolls, E. T., Kingston, B., Megson,A., & Gunary, R. (1981). Variety in a meal enhancesfood intake in man. Physiology & Behavior, 26, 215–221.

Rolls, B. J., Wood, R. J., & Rolls, E. T. (1980). Thirst:The intiation, maintenance, and termination ofdrinking. In J. M. Sprague & A. N. Epstein (Eds.),Progress in psychobiology and physiological psychology (pp.263–321). New York: Academic Press.

Rybiski, L. R. (1996). A test for stimulus-specificity in thewithin-session pattern of responding. Unpublished mas-ter’s thesis, McNeese State University, Lake Charles,LA.

Savory, C. J. (1988). Rates of eating in domestic fowls inrelation to changing food deficits. Appetite, 10, 57–65.

Schoenfeld, W. N., Antonitis, J. J., & Bersh, P. J. (1950).Unconditioned response rate of the white rat in a bar-pressing apparatus. Journal of Comparative & Physiolog-ical Psychology, 43, 41–48.

Sokolov, Y. N. (1963). Perception and the conditioned reflex(S. W. Waydenfeld, Trans.). Oxford, England: Per-magon Press.

Staddon, J. E. R. (1967). Attention and temporal dis-crimination: Factors controlling responding under acyclic-interval schedule. Journal of the ExperimentalAnalysis of Behavior, 10, 349–359.

Staddon, J. E. R., & Higa, J. J. (1996). Multiple timescales in simple habituation. Psychological Review, 103,720–733.

Susser, M. (1973). Causal thinking in the health sciences:Concepts and strategies of epidemiology. New York: OxfordUniversity Press.

Swithers, S. E., & Hall, W. G. (1994). Does oral experi-ence terminate ingestion? Appetite, 23, 113–138.

Teyler, T. J., Chiana, N., DiScenna, P., & Roemer, R. A.(1984). Habituation of central nervous system evokedpotentials: Intrinsic habituation examined in neocor-tex, allocortex and mesencephalon. In H. V. S. Peeke& L. Petrinovich (Eds.), Habituation, sensitization andbehavior (pp. 251–285). New York: Academic Press.

Thompson, R. F., & Spencer, W. A. (1966). Habituation:A model phenomenon for the study of neuronal sub-strates of behavior. Psychological Review, 73, 16–43.

Thorpe, W. H. (1966). Learning and instinct in animals.Cambridge, MA: Harvard University Press.

Treit, D., Spetch, M. L., & Deutsch, J. A. (1983). Varietyin the flavor of food enhances eating in the rat: Acontrolled demonstration. Physiology & Behavior, 30,207–211.

Wagner, A. R. (1976). Priming in STM: An information-processing mechanism for self-generated or retrieval-generated depression in performance. In T. J. Tighe& R. N. Leaton (Eds.), Habituation: Perspectives fromchild development, animal behavior and neurophysiology(pp. 95–128). New York: Wiley.

Weatherly, J. N., McSweeney, F. K., & Swindell, S. (1995).On the contributions of responding and reinforce-ment to within-session patterns of responding. Learn-ing and Motivation, 26, 421–432.

Weatherly, J. N., McSweeney, F. K., & Swindell, S. (1997).Within-session patterns of responding with changes inthe variability and probability of food delivery. Behav-ioural Processes, 39, 279–289.

Weingarten, H. P. (1985). Stimulus control of eating:Implications for a two-factor theory of hunger. Appe-tite, 6, 387–401.

Whitlow, J. W., Jr. (1975). Short-term memory in habit-uation and dishabituation. Journal of Experimental Psy-chology: Animal Behavior Processes, 1, 189–206.

Wyers, E. J., Peeke, H. V. S., & Herz, M. J. (1973). Be-havioral habituation in invertebrates. In H. V. S.Peeke & M. J. Herz (Eds.), Habituation: Vol. 1. Behav-ioral studies (pp. 1–57). New York: Academic Press.

Received May 17, 2000Final acceptance September 25, 2000

362

ERRATUM

Kennedy, C. H., Meyer, K. A., Werts, M. G., & Cushing, L. R. (2000). Effectsof sleep deprivation on free-operant avoidance. Journal of the Experimental Analysisof Behavior, 73, 333–345.

Error: Citation of the effects of sleep deprivation in Kennedy and Itkonen(1993) and Iwata (1994). Correction: Neither article refers to sleep deprivation.

criticisms of the satiety hypothesis as an explanation for within-session decreases in responding

Documents