Salt-induced thirst results in increased finickiness in humans.
Abstract: Common sense suggests that water-deprived or food-deprived organisms should be more willing to consume foods or fluids that would be deemed undesirable under lower states of deprivation. With food, evidence favoring this account has been observed; however other studies find that hungry participants demonstrate increased finickiness--avoiding less palatable alternatives and consuming more of pleasant ones. This study set out to test whether thirst generally increases acceptability (the "common sense" view) or whether it produces increased finickiness. Thirst was induced in one group of participants (relative to nonthirsty controls), and the impact of this on their judgment of optimal and suboptimal water sources was examined. Thirsty participants liked optimal stimuli more, but liked suboptimal stimuli less, relative to controls, but this occurred only when the stimuli were actually tasted. These data suggest that thirst polarizes extant hedonic responses to fluids--finickiness--thereby maintaining optimal water selection under conditions of mild to moderate deprivation.

Key words: thirst, palatability, finickiness, humans
Article Type: Report
Subject: Salt (Physiological aspects)
Salt (Psychological aspects)
Stimuli (Psychology) (Physiological aspects)
Thirst (Psychological aspects)
Authors: Stevenson, Richard J.
Case, Trevor I.
Oaten, Megan J.
Pub Date: 06/22/2010
Publication: Name: The Psychological Record Publisher: The Psychological Record Audience: Academic Format: Magazine/Journal Subject: Psychology and mental health Copyright: COPYRIGHT 2010 The Psychological Record ISSN: 0033-2933
Issue: Date: Summer, 2010 Source Volume: 60 Source Issue: 3
Product: SIC Code: 2899 Chemical preparations, not elsewhere classified
Geographic: Geographic Scope: Australia Geographic Code: 8AUST Australia
Accession Number: 233492508
Full Text: Basic motivational states such as thirst and hunger act to energize and organize behavior toward satisfying these specific needs. Such motivational states also make cues that signal drink and food more salient and desirable, as well as affecting hedonic reaction to the particular goal stimulus. Food has been relatively well studied with respect to the latter. A particular food may be judged as more pleasant at the beginning of a meal than at the end (e.g., Rolls & Rolls, 1997), and food in general--its flavor, sight and smell--may be less appealing for some time after a meal than in the hour or so before the next (e.g., Duclaux, Feisthauer, & Cabanac, 1973). These affective reactions presumably act to temper an individual's consumption, such that greater need, be it thirst or hunger, is typically associated with a more positive hedonic response to drink or food than when he or she is sated (Sorensen, Moller, Flint, Martens, & Raben, 2003). For many animals, and indeed for a significant number of humans, too, there are occasions when optimal sources of water and food are unavailable. Water may be warm or muddy and food bland, insect infested, or stale. Typically, such water and food sources would be avoided if more positively valenced items were available--cool, transparent water and high-energy food. The issue explored here concerns how motivational state affects hedonic responding to salient goal objects when those objects vary in palatability. In particular, the study sought to investigate whether thirst acts to make any satiating water source more pleasant (or less unpleasant), or whether it acts in a different manner.

How thirsty humans react to water sources that vary in palatability has not been investigated, although it is known that thirst increases positive hedonic evaluations of room-temperature tapwater (Rolls et al., 1980). Anecdotally, it might be suspected that in high-need states, water would be greeted with considerable enthusiasm, and this does seem to be the case. Sven Hedin, a 19th-century explorer, recounted his experience of encountering water following a protracted period of deprivation in the Taklamakan desert:

However, whether thirst results in more positive (or less negative) hedonic reactions to unpalatable fluids is a different matter altogether. Again, anecdotal reports suggest that, in extremis, such fluids are drunk but that they are not accompanied by any lessening of negative affect. For example, take the case of the journalist Ryszard Kapuscinski, who hitched a ride on a local truck in the Sahara. The truck broke down and, though almost delusional with thirst, Kapuscinski (2008) reported that the water available (muddy, gritty water stored in a poorly cured goat skin) was disgusting. So, on this basis it might be suggested that though normally, fresh, pleasant water becomes even more enjoyable if one is thirsty, the same may not apply to water that would, in a nondeprived state, be regarded as undrinkable.

Although anecdotes may provide some insights, it is important to ask whether there is any evidence from other motivational domains. For food, two findings suggest that deprived or chronically hungry participants may be more accepting of less palatable food than nondeprived participants. Hoefling et al. (2009) found that hungry participants showed less disgust, as measured by facial electromyography (EMG) on the levator muscle region, in response to pictures of unpalatable foods, relative to controls. In a very different study, but with a conceptually similar outcome, Dykens (2000) examined self-report food choice in a group of participants with Prader-Willi syndrome, compared to controls broadly matched for IQ. Prader-Willi syndrome is characterized by an incessant desire to eat--chronic hunger--and such individuals often become morbidly obese. Dykens found that the participants with Prader-Willi syndrome had intact knowledge relating to food safety and appropriateness, yet they reported a willingness to eat food combinations and food sources that were deemed unacceptable by the controls. Dykens concluded that "their drive for food is so strong that it overpowers their knowledge and otherwise good judgment" (p. 164). Whereas these findings seem to hold with the view that deprivation results in a greater willingness to consume unpalatable foods, many other studies have come to a rather different conclusion.

Schachter (1971) suggested that food-relevant cues (e.g., taste, smell) affect consumption to a greater degree when the person (or animal) is hungry. This perspective leads to a rather different and contraintuitive prediction, namely, that hungry subjects will consume less of an unpalatable food than nonhungry participants. This type of outcome has in fact been observed in both humans and animals. Kauffman, Herman, and Polivy (1995) gave hungry or nonhungry participants access to either a palatable milkshake or an unpalatable quinine-adulterated milkshake. The hungry participants consumed significantly less of the unpalatable milkshake relative to the non-hungry individuals. Similarly, Jacobs and Sharma (1969) observed that hungry dogs and rats rejected disliked tastes more than nonhungry controls. This type of behavior, namely, rejection of motivationally relevant but less palatable items under conditions of deprivation, has been termed finickiness (Pliner, Herman, & Polivy, 1990).

With thirst, the consequences of deprivation are less clear, as there are no human studies exploring how thirst affects responses to fluids that systematically vary in palatability. In rats, thirst appears to generate finickiness under conditions of uncontrollable stress (Dess, Chapman, & Minor, 1988; Minor, Dess, Ben-David, & Chang, 1994) and when the animal's immune system is activated by exposure to bacterial proteins (Aubert & Danzer, 2005). Whether thirst induces finickiness in animals or greater acceptability of both palatable and unpalatable fluids does not appear to have been addressed, although the anecdotal report of Kapuscinski in the Sahara might suggest the former. The aim of the experiment reported here was to determine whether both positively and negatively valenced fluids would be liked by thirsty participants or whether finickiness would emerge (i.e., positive liked more, negative liked less) relative to nonthirsty controls.

To induce thirst, a slightly modified form of the manipulation developed by Changizi and Hall (2001) was adopted, which involved participants' consuming salted potato crisps. This manipulation induces a strong and reliable thirst, and participants receiving it were then compared to participants fed unsalted crisps with ad lib access to water (Calorie control) and to a further group just given ad lib access to water (No calorie control). To assess the effect of thirst on affective reactions to fluid stimuli varying in palatability, four different types of test were used. First, small samples of water were tasted at different temperatures, because increasing temperature is known to reduce liking for plain water (Zellner, Stewart, Rozin, & Brown, 1988). Second, a series of quinine solutions that varied in bitterness were sampled, from no bitter agent present to moderate levels of bitterness. Bitter tastes are reportedly unpleasant to both humans and animals (Glendinning, 1994). Third, participants then viewed sampled water containing either odorless and tasteless food colorants or plain water. Fourth, participants were shown a series of pictures of various water sources, which varied in affective valence. As the primary interest was in affective reaction, participants were asked in each case to judge their hedonic reaction to the stimulus (Bauer & Altarriba, 2008), their degree of disgust, and the extent to which the fluid was refreshing.


Sixty healthy participants (mean age = 23.7, SD = 7.0), recruited from the Macquarie University community, took part for a small cash payment. Participants were randomly assigned to three experimental groups, with approximately equal numbers of men and women within each group. As sex had no significant impact on the variables studied here, it is not further reported in the manuscript.


Participants in the Thirsty condition consumed 50 g of salted potato crisps (Smiths Original) containing 0.75 g of salt, with an energy value of 1065 kJ. Participants in the Calorie control condition consumed 50 g of unsalted potato crisps (Freedom Foods) containing 0.04 g of salt, with an energy value of 1140 kJ.

The temperature task used tap water presented at 2, 22 (room temperature), 40, and 60C. Stimuli were prepared immediately before use from thermos flasks and were presented in 10-ml aliquots in clear plastic sample cups. The taste task used tap water presented at room temperature, either with no quinine added or with quinine (Sigma) at a concentration of 0.0002%, 0.001%, or 0.003%. These were presented in the same manner as for the temperature task. The color task used either uncolored room temperature tap water or water colored light gray or dark gray. The dyes used to color the water appeared to have no detectable taste or smell in a prior pilot experiment (see Stevenson, Case, & Tomiczek, 2007, for olfactory testing methods). These too were presented to participants in the same manner as for the temperature task.

Images used in the picture task were drawn from a variety of sources (including the International Affective Picture Series; Lang, Bradley, & Cuthbert, 2001) and were presented to participants as A4-size color copies.


Participants completed the study alone with the experimenter, between 11 a.m. and 5 p.m. (because earlier in the day was not conducive to consuming crisps). On arrival, participants were randomly assigned to one of three experimental groups, Thirst, Calorie control, or No calorie control. After obtaining consent (the study was approved by the Macquarie University Human Research Ethics Committee), participants in the Thirsty group were asked to eat their salted crisps. After they had finished eating, they were asked to read magazines for 15 min so that they were suitably thirsty when the experiment proper began. Participants in the Calorie control condition completed a similar procedure, except that they were given low-salt crisps. Importantly, they were then instructed to drink 200 ml of water and in addition were allowed ad lib access to water in the 15 min following crisp consumption (mean extra amount drunk in ml = 106.5, SD = 106.9). Participants in the No calorie control condition were also instructed to drink 200 ml of water and to read magazines for 15 min. They too had ad lib access to additional water in this period (mean extra amount drunk in ml = 56.5, SD = 90.3). After the 15 min had elapsed, all participants evaluated how hungry, thirsty, happy, sad, calm, and excited they were on 9-point category scales (1 = not at all, 9 = very).

Participants in all three groups then completed the four test modules. The modules were presented in counterbalanced order across participants within each group (i.e., the first participant in the Thirsty group received the same presentation order as the first participant in the Calorie control and No calorie control groups, etc). The four modules were composed of the temperature test, the taste test, the color test, and the picture test. Stimulus order within each test (excepting the picture test, which alternated between two fixed randomized orders) was also counterbalanced across participants within each group.

Temperature test. The participants were asked to pour the entire stimulus into their mouths and then expectorate. They were then asked to evaluate the stimulus on three 9-point category scales, for liking/disliking (1 = like extremely, 9 = dislike extremely), how disgusted they felt (1 = not at all, 9 = very), and how refreshing they found the water (1 = not at all, 9 = very). A 30-sec interstimulus interval separated each presentation.

Taste test. This was identical to the temperature test, except here participants were asked to rinse between trials to minimize taste carryover.

Color test. This was identical to the temperature test, except here participants were asked to evaluate how much they would like to drink a whole glass of the stimulus prior to actually sampling it. This too was rated on a 9-point category scale (1 = not at all, 9 = very much).

Picture test. On this test participants were shown 18 pictures, each of which had a label describing the picture to ensure that participants were aware of its salient features. The descriptive notes that follow were included with the pictures; the notes in parentheses are only given here, to clarify the nature of each image.

1. This is a glass of tap water.

2. This is water in a backyard swimming pool.

3. This is a beef and rice dish. An unidentified hair rests in the middle of the dish.

4. These are a pair of socks that have been worn repeatedly for a week without washing.

5. This is water in a brand new dog's drinking bowl.

6. This is pasta with a pasta sauce.

7. This is a fresh water stream (muddy water).

8. This is a piece of cake with a dead fly resting in the center.

9. This is water in a brand new bedpan.

10. This is a pile of a stranger's clothing.

11. This is a set of rotting teeth.

12. This is water coming out of a public tap. (Third World setting)

13. This is a stranger's half-eaten meal.

14. This is a secondhand but fully laundered unisex t-shirt.

15. This is a sponge that has been used to clean the toilet, shower, and bathroom sink.

16. This is a used facial tissue.

17. This is a plate of fried grasshoppers.

18. This is a whole pig on a spit roast.

Each stimulus was evaluated using a picture-specific, 5-point contact hierarchy scale, running from not even wanting to look at the image (scored 1) through to some form of oral contact (scored 5).

After the four test modules were finished, participants then completed the same set of rating scales that preceded these tests (i.e., rating hunger, thirst, etc.). All participants then filled in the disgust sensitivity scale (Haidt, McCauley, & Rozin, 1994), to check for differences between groups.


Table 1 presents mean thirst, hunger, happy, sad, calm, and excited ratings obtained following the thirst induction, and again at the end of the experiment. Thirst ratings were examined first to ensure that the manipulation had worked. A two-way ANOVA, with Group (Thirsty vs. Calorie control vs. No calorie control) as a between-participants factor and Time (Initial vs. Final measure) as a within-participant factor, revealed only one significant effect, that of Group, F(2, 57) = 15.14, MSE = 5.42, p < 0.001, eta-squared = 0.35. As is evident in Table 1, and as confirmed by post hoc Bonferroni adjusted contrasts, thirst ratings were significantly higher in the Thirsty group, relative to the two control conditions (ps < 0.001). The two control conditions did not significantly differ. All of the other rating types were also analyzed, including the disgust sensitivity scores. No significant differences involving Group were found, excepting a main effect of Group for hunger ratings; however, post hoc contrasts on this main effect failed to reveal any differences. As there were no significant differences between the two control groups on any of the ratings here--or, indeed, for any of the four tests reported below--the two control conditions were combined (i.e., treated as one single condition) and are reported and analyzed as a single control group in what follows.

Temperature Test

All three evaluations--hedonic, disgust, and refreshing (see Table 2 for means)--were significantly correlated for each sample (range of absolute rs = 0.27-0.60). Consequently, after reverse coding of the refreshing ratings, the three measures were averaged for each stimulus. This composite evaluative scale was then entered into a two-way ANOVA, with Group (Thirsty vs. Control) as a between-participants factor and Temperature (2C, 22C, 40C, 60C) as a within-participant factor. The analysis revealed two significant effects. First, as expected, evaluative ratings became increasingly negative with increasing water temperature, F(3, 174) = 47.54, MSE = 1.39, p < 0.001, eta-squared = 0.45, confirming the utility of this independent variable on evaluative responding. Second, there was a Group x Temperature interaction, F(3, 174) = 3.61, MSE = 1.39, p < 0.02, eta-squared = 0.06. Examination of Figure 1 suggests a flatter set of evaluations in the Control group, relative to the Thirsty group, and this interpretation is borne out by the significant linear trend interaction of Temperature x Group, F(1, 58) = 6.72, MSE - 2.18, p < 0.02, eta-squared = 0.10. This effect suggests more positive responses to cooler water and more negative responses to warmer water in thirsty participants relative to controls.


Taste Test

All three evaluations (see Table 3 for means) were again significantly related for each sample (range absolute of rs = 0.39--0.70). These data were collapsed as described for the temperature test and then subjected to a two-way ANOVA, with Group (Thirsty vs. Control) as a between-participants factor and Taste (bitterness level 0 vs. 1 vs. 2 vs. 3) as a within-participant factor. Two significant effects emerged. First, as expected, evaluative ratings became increasingly negative with increasing quinine concentration, F(3, 174) = 117.23, MSE = 1.54, p < 0.005, eta-squared = 0.67. Second, there was a Group x Taste interaction, F(3, 174) = 5.15, MSE = 1.54, p < 0.005, eta-squared = 0.08. Examination of Figure 2 suggests that controls reported more negative evaluations as bitterness increased, whereas those in the Thirsty group judged the room-temperature water blank in a manner equivalent to controls, although being more tolerant to the lowest concentration of quinine and less tolerant of the two higher concentrations. This interpretation was buttressed by the significant cubic trend interaction between Group and Taste, F(1, 58) = 14.05, MSE = 1.79, p < 0.001, eta-squared = 0.17.


Color Test

Desire-to-drink ratings, which were solely based on appearance (see Table 4) were analyzed first. Using a two-way ANOVA, with Group (Thirsty vs. Control) as a between-participants factor and Color (none vs. gray vs. dark) as a within-participant factor, revealed just one significant effect--Color. Participants reported a greater desire to drink the transparent sample, relative to the colored fluids, F(2, 116) = 46.56, MSE = 2.45, p < 0.001, eta-squared = 0.44.

All three evaluations (see Table 4 for means) were again significantly related for each sample (range of absolute rs = 0.27-0.71). These data were collapsed as described for the temperature test and then subjected to a two-way ANOVA (as for taste and temperature). There was a main effect of Color, F(2, 116) = 22.21, MSE = 0.87, p < 0.001, eta-squared = 0.27), and the interaction of Color and Group (F = 2.77) approached significance (p = 0.067). Participants in the Thirsty group evaluated uncolored water more positively and colored water less positively than participants in the Control group (see Figure 3). Using a post hoc t test, the difference in evaluative response between the uncolored stimulus and the two colored stimuli (averaged) between the Thirsty and Control groups was examined. The difference in evaluative response between the clear and colored fluids was significantly larger in the Thirsty group, t(58) = 2.14, p < 0.05, [r.sup.2] = 0.07, relative to the control group.


Aggregate Analysis of the Three Tasting Tests

For the three tests above that involved tasting actual stimuli--temperature, taste and color--a mean overall evaluative rating for palatable stimuli (i.e., cold water, water without quinine, and uncolored water [three scores]) and for unpalatable stimuli (i.e., warmer waters, water with quinine, and colored waters [eight scores]) was calculated. Neither of these mean overall scores differed by group (ts = 0.8 and 1.1, respectively); however, the two scores did significantly differ by group, t(58) = 2.32, p < 0.025, [r.sup.2] = 0.08). In thirsty participants, the difference in evaluative response between palatable (M= 3.4) and unpalatable (M= 5.5) fluids was greater (Mdiff = -2.1) than that for nonthirsty controls (palatable M = 3.7; unpalatable M = 5.2; Mdiff = -1.5). This result, taken in conjunction with the findings above, is highly consistent with increased finickiness in thirsty subjects.

Picture Test

The picture stimuli were coded so that six scores were assembled for each participant: pleasant and unpleasant drink-related scores, food-related scores, and non-food- or drink-related scores (see Table 5 for means). These were then entered into a three-way ANOVA, with Group (Thirsty vs. Control) as a between-participants factor and picture Valence (pleasant vs. unpleasant) and picture Type (drink vs. food vs. non-food/drink) as within-participant factors. The ANOVA revealed significant main effects for Valence, F(l, 58) = 688.03, MSE = 0.44, p < 0.001, eta-squared = 0.92, and Stimulus type, F(2, 116) = 147.63, MSE = 0.40, p < 0.001, eta-squared = 0.72, and an interaction between these factors, F(2, 116) = 16.59, MSE = 0.49, p < 0.001, eta-squared = 0.22. However, while this indicates the broadly expected differences in response (but noting a larger disparity between positively and negatively valenced food images, relative to water and non-food or drink images), no significant effects involving Group were found. Thus, on this test, thirsty and control participants did not differ in their evaluations.


This experiment set out to explore how participants, thirsty or sated, responded to fluids that differed in palatability. Based on the limited literature available on thirst and the somewhat more extensive one on food, the expectation was that one of two outcomes would occur: Thirsty participants would be more positive in their evaluation of both palatable and unpalatable fluids, or thirsty participants would demonstrate finickiness, with more positive evaluations of palatable fluids and less positive evaluations of unpalatable fluids, relative to non-thirsty controls. For tests involving direct contact with fluids, namely temperature, taste, and the second part of the color task, interaction effects were evident and the pattern of results in each case suggested that more palatable stimuli were evaluated more positively and less palatable stimuli less positively by thirsty subjects than by nonthirsty controls. A further test was conducted pooling all of the palatable stimuli and contrasting this to all of the unpalatable stimuli across the three fluid contact tests. The result clearly favored increased finickiness, as thirsty participants' evaluations were significantly more polarized than nonthirsty participants'. Indeed, across all of the tasting tests, there was no evidence favoring a general increase in palatability to all fluids in thirsty participants. Thus, salt-induced thirst seems to exaggerate evaluative reactions present in nonthirsty controls. The exception to this conclusion seems to arise when participants view, rather than taste, fluid samples. On the picture task, images of water of varying palatability were evaluated in a similar manner by the thirsty and control participants, suggesting that differences in reported affective reactions depend on the participants' actively sampling the test solution. Similarly, as with the color test, a priori evaluations of the colored fluids based on their appearance did not differ by group. Interestingly, both of the studies reported in the introduction that observed greater tolerance toward unpalatable foods in either food-deprived participants (Hoefling et al., 2009) or chronically hungry participants (Dykens, 2000) involved either pictures or vignettes. In contrast, studies reporting finickiness utilized real food stimuli (Jacobs & Sharma, 1969; Kauffman et al., 1995).

When thirsty participants sample real fluids of varying palatability, they seem to respond in a manner similar to the anecdotal example considered in the introduction--foul water is still foul water even when thirsty, and, possibly, thirsty individuals perceive foul-tasting water as being even worse than do nonthirsty controls. This conclusion seems contraintuitive, as one might expect that an animal or person in a deprived state would be less choosy than in a nondeprived state. Several possible explanations have been advanced for this type of finding in the food literature, and before considering how these might apply to the findings reported here in relation to thirst, it is important to consider whether these results might have arisen from some more mundane cause. One possible explanation is that the thirsty group differed in the way in which they used the rating scales. This would appear to be unlikely for two reasons. First, group assignment was random, so systematic differences in scale usage would not be expected. Second, hedonic evaluations on the picture test did not differ between groups, as one might expect if the thirsty participants were using the scales differently. Thus, scale use differences seem to be an unlikely explanation of the observed group differences. A further possibility is that participants in the nonthirst conditions might simply have been demotivated and thus less affectively responsive. This, too, would seem unlikely, as emotion/arousal self-reports did not differ between groups, and all groups reported the warm water and stronger quinine solutions as unpleasant.

As noted in the introduction, Schachter (1971) suggested that organisms become more sensitive to external sensory cues under conditions of deprivation (and see Valdivia, Luciano & Molina, 2006). Similarly, Kauffman et al. (1995) noted that other motivational theorists have also suggested that higher drive states might be associated with increasing differential responsiveness to environmental stimuli (e.g., Easterbrook, 1959). This type of account suggests that as a particular deprivation state increases, the organism pays more attention to sensory information that relates to that state. Thus, tastes, color, and temperature would all constitute relevant fluid-related cues, with thirsty participants becoming more discriminating (rather than less) about what might or might not be consumed (the assumption here being that evaluative ratings broadly reflect how much of the fluid would actually be consumed). Although this account is consistent with our findings, it does seem to lack a functional rationale. After all, if an organism was very thirsty, one would expect that it would be better to drink and risk gastrointestinal illness or poisoning than death via dehydration. Although this might indeed be the case under conditions of extreme deprivation (and anecdotal accounts of food consumption in extremis are certainly consistent with this; see Kauffman et al., 1995), thirst in the low to moderately high range may be under different functional constraints. This may result from the probability of these states occurring in the natural environment. Low-to-moderate-range deprivation is likely to be fairly frequently experienced by many organisms, including a still substantial part of the human race. Rather than risking illness from contaminated water or poisoning from bitter-tasting fluids, it may be better to search for a more palatable source, because the chance of finding one probably exceeds the risk associated with not finding one. More importantly, the risk of consuming suboptimal fluids in the context of the probability of finding a more optimal water source makes it functionally useful to delay consumption by becoming more finicky about what is drunk.


AUBERT, A., & DANZER, R. (2005). The taste of sickness: Lipopolysaccharide-induced finickiness in rats. Physiology and Behavior, 84, 437-444.

BAUER, L. M., & ALTARRIBA, J. (2008). An investigation of sex differences in word ratings across concrete, abstract and emotion words. The Psychological Record, 58, 465-474.

CHANGIZI, M. A., & HALL, W. G. (2001). Thirst modulates a perception. Perception, 30, 1489-1497.

DESS, N. K., CHAPMAN, C. D., & MINOR, T. R. (1988). Inescapable shock increases finickiness about drinking quinine-adulterated water in rats. Learning and Motivation, 19, 408-424.

DUCLAUX, R., FEISTHAUER, J., & CABANAC, M. (1973). Effets du repas sur l'agrement d'odeurs alimentaires et nonalimentaires chez l'homme. Physiology and Behavior, 10, 1029-1033.

DYKENS, E. M. (2000). Contaminated and unusual food combinations: What do people with Prader-Willi syndrome choose? Mental Retardation, 38, 163-171.

EASTERBROOK, J. A. (1959). The effect of emotion on cue utilization and the organization of behavior. Psychological Review, 66, 183-201.

GLENDINNING, J. I. (1994). Is the bitter rejection response always adaptive? Physiology and Behavior, 56, 1217-1227.

HAIDT, J., MCCAULEY, C., & ROZIN, P. (1994). Individual differences in sensitivity to disgust: A scale sampling seven domains of disgust elicitors. Personality and Individual Differences, 16, 701-713.

HOEFLING, A., LIKOWSKI, K. U., HAFNER, M., SEIBT, B., MUHLBERGER, A., WEYERS, P., & STRACK, F. (2009). When hunger finds no fault with moldy corn: Food deprivation reduces food-related disgust. Emotion, 9, 50-58.

JACOBS, H. L., & SHARMA, K. N. (1969). Taste versus calories: Sensory and metabolic signals in the control of food intake. Annals of the New York Academy of Sciences, 157, 1084-1125.

KAPUSCINSKI, R. (2008). Shadow of the sun. Melbourne, Australia: Penguin.

KAUFFMAN, N. A., HERMAN, P., & POLIVY, J. (1995). Hunger-induced finickiness in humans. Appetite, 24, 203-218.

LANG, P. J., BRADLEY, M. M., & CUTHBERT, B. N. (2001). International affective picture system (IAPS): Instruction manual and affective ratings (Tech. Rep. A-5). Miami: The Center for Research in Psychophysiology, University of Florida.

MINOR, T. R., DESS, N. K., BEN-DAVID, E., & CHANG, W. (1994). Individual differences in vulnerability to inescapable shock in rats. Journal of Experimental Psychology: Animal Behavior Processes, 20, 402-412.

PLINER, P., HERMAN, C. P., & POLIVY, J. (1990). Palatability as a determinant of eating: Finickiness as a function of taste, hunger and the prospect of good food. In E. D. Capaldi & T. L. Powley (Eds.), Taste, experience and feeding (pp. 210-226). Washington, DC: APA.

ROLLS, B. J., & ROLLS, E. T. (1982). Thirst. Cambridge, UK: Cambridge University Press.

Rolls, B. J., Woods, R. J., Rolls, E. T., Lind, H, Lind, W., & Ledingham, J. G. G. (1980). Thirst following water deprivation in humans. American Journal of Physiology, 23, R476-482.

ROLLS, E. T., & ROLLS, J. H. (1997). Olfactory sensory-specific satiety in humans. Physiology and Behavior, 61, 461-473.

SCHACHTER, S. (1971). Some extraordinary facts about obese humans and rats. American Psychologist, 26, 129-145.

SORENSEN, L. B., MOLLER, P., FLINT, A., MARTENS, M., & RABEN, A. (2003). Effect of sensory perception of foods on appetite and food intake: A review of studies on humans. International Journal of Obesity, 27, 1152-1166.

STEVENSON, R. J., CASE, T. I. & TOMICZEK, C. (2007). Resistance to interference of olfactory perceptual learning. The Psychological Record, 57, 103-116.

VALDIVIA, S., LUCIANO, C., & MOLINA, F. J. (2006). Verbal regulation of motivational states. The Psychological Record, 56, 577-595.

ZELLNER, D. A., STEWART, W. F., ROZIN, P., & BROWN, J. M. (1988). Effect of temperature and expectations on liking for beverages. Physiology and Behavior, 44, 61-68.

Richard J. Stevenson, Trevor I. Case, and Megan J. Oaten

Macquarie University, Sydney, Australia

Correspondence should be addressed to Richard J. Stevenson at the Department of Psychology, Macquarie University, Sydney, NSW2109, Australia (e-mail:
I stood on the brink of a little pool filled with fresh water. It
  would be in vain for me to describe the feelings which overcame me.
  I took my tin box, filled it and drank. How sweet that water tasted!
  Nobody can conceive it who has not been within an ace of dying of
  thirst. I lifted the tin to my lips and drank, drank, drank, time after
  time. How delicious! (cited in Rolls & Rolls, 1982, p. 32).

Table 1

Mean (and SD) Rating Scores Obtained Near the Beginning (First) and
Toward the End (Second) of the Experiment

Group                              Evaluation


                            Thirst     Hunger    Happy

                             M   SD    M   SD    M    SD

                    First   7.3  1.7  3.9  1.9  6.6  1.3
                    Second  6.6  1.9  3.9  2.0  6.6  1.3

Calorie control
                    First   4.3  1.5  3.6  1.6  6.6  1.4
                    Second  4.5  2.3  4.3  2.1  6.4  1.4

No calorie control
                    First   4.4  1.8  2.7  1.8  6.3  1.4
                    Second  4.7  2.5  2.7  2.1  6.5  1.5

Group                              Evaluation


                              Sad      Calm     Excited

                             M   SD    M   SD    M    SD

                    First   2.5  1.6  7.6  1.5  4.1  1.9
                    Second  2.4  1.7  7.3  1.5  4.8  2.2

Calorie control
                    First   2.8  1.6  7.1  1.7  4.9  2.1
                    Second  3.0  1.7  7.0  1.4  4.4  1.7

No calorie control
                    First   2.4  1.7  7.4  1.4  4.1  1.7
                    Second  2.2  1.6  7.0  1.5  4.7  2.3

Table 2

Mean Hedonic, Disgust and Refreshing Ratings (and SDs) for the Water
Stimuli When Presented at Different Temperatures to the Thirsty and
Control Groups

Group                       Water temperature


                        2C       22C       40C       60C

                      M   SD    M   SD    M   SD    M   SD

         Hedonic     2.4  1.6  3.5  1.7  4.7  1.8  5.2  2.1
         Disgust     1.3  0.9  2.1  1.7  2.9  2.0  3.3  2.2
         Refreshing  7.6  2.3  5.8  2.3  3.6  2.1  3.1  2.0

         Hedonic     3.6  1.4  4.1  1.4  4.9  1.5  4.9  1.5
         Disgust     1.9  1.4  2.1  1.7  2.7  1.6  2.6  1.8
         Refreshing  6.7  2.2  5.8  2.0  3.7  2.1  3.9  2.0

Table 3

Mean Hedonic, Disgust and Refreshing Ratings (and SDs) for the Bitter
Stimuli When Presented to the Thirsty and Control Groups

Group                        Quinine concentration


                       0.0%     0.0002%   0.001%    0.003%

                      M   SD    M   SD    M   SD    M   SD

         Hedonic     4.5  1.8  4.4  2.3  7.9  1.2  8.4  1.0
         Disgust     4.1  1.9  3.4  2.1  7.5  1.4  8.1  1.4
         Refreshing  4.6  2.6  4.7  2.7  1.9  1.1  1.3  0.7

         Hedonic     4.7  1.7  5.5  1.9  7.2  1.5  7.9  1.3
         Disgust     3.0  2.2  4.2  2.5  6.2  2.3  7.6  1.4
         Refreshing  5.0  2.0  4.0  1.9  2.7  1.9  1.9  1.4

Table 4

Mean Desire, Hedonic, Disgust and Refreshing Ratings (and SDs) for the
Colored Stimuli When Presented to the Thirsty and Control Groups

Group                          Stimulus


                     Uncolored  Light gray  Dark gray

                      M   SD      M   SD     M   SD

         Desire      6.3  1.9    3.8  2.0   3.8  2.2
         Hedonic     3.6  2.0    5.1  1.8   5.3  1.7
         Disgust     2.8  2.2    4.0  2.1   4.1  2.2
         Refreshing  5.5  2.4    4.6  2.0   3.9  2.0

         Desire      6.3  1.7    4.0  1.7   3.2  1.6
         Hedonic     4.4  1.6    4.6  1.6   5.0  1.6
         Disgust     2.9  2.1    3.2  2.0   3.7  2.3
         Refreshing  5.5  2.0    4.9  1.8   4.6  1.9

Table 5

Mean Approach Ratings (and SDs) for the Pictorial Stimuli for the
Thirsty and Control Groups Organized by Picture Type and Valence

Group                    Picture type and valence

               Drink                Food               Other

         Positive  Negative  Positive  Negative  Positive  Negative

          M   SD    M   SD    M   SD    M   SD    M   SD    M   SD

Thirsty  4.7  0.7  2.8  0.6  5.0  0.0  2.2  0.8  2.9  0.9  1.6  0.4
Control  4.7  0.8  2.9  0.6  4.7  0.8  2.5  0.9  3.4  1.0  1.8  0.6
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.