Music as a discrimination in rats: An analysis of stimulus generalization

Jennifer Reddish
 JenLR11@aol.com

                                                 Saint Anselm College Psychology Department Home Page
 

Acknowledgements
Results
Abstract
Discussion
Introduction
References
Methods
Appendices

 
 





Acknowledgements

    First and foremost I would like to thank Bernie, Phil, Number 2, Lola, Tamiqua, Jones, Lucky, and Iverson.  They showed up every day ready to go, and without them this study could never have been accomplished, Thanks Guys!  Thanks to Professor Troisi for his patience, and for the fact that he always took the time, even when he was extremely busy, to make sure that I was on the right track with my study.  I really appreciate it!  Thank you also to Professor Flannery who helped me with my study.  I would also like to thank Carson, Amy and Katie for being there to talk to about my thesis and still being there when I was a bit emotional about my rats. I have one word for you girls……… BAAALTIMORE!  Ry, thanks for all the “don’t stress, life is to short to worry” talks, they really did help me relax and not stress out (well not stress out a lot).  Mom, you’re the best, Love you and Pat!  Ahhhhhhh TGIF (Thank Goodness Its Finished)!

Back to top

Abstract

    Although a large amount of research has been conducted on animal discrimination tasks, research findings using rat’s response to audio stimuli has been studied less frequently.   The purpose of this study was to examine whether eight previously trained rats would have a greater response rate of target behavior when stimulated by a piece of music in the allegro tempo in the major mode than when they were stimulated by a piece of music in adagio tempo in the minor mode.  It was predicted that a distinction would be made between whether (1) subjects would be able to discriminate between the two complex stimuli, and whether (2) subjects would have a greater response rate for target behavior when stimulated by the allegro tempo piece of music in major mode.  8 subjects were divided into two groups of 4 and were assigned to separate two-lever operant chambers.  These chambers were defined as Box A (left lever pressing was reinforced by a food pellet) and Box B (right lever pressing was reinforced by a food pellet).  All subjects were assigned to Condition A (Samuel Barber’s Adagio for Strings,) and were assigned to either Box A or Box B.  All subjects were then assigned to Condition B (Rachmaninoff’s Rach Second Symphony Allegro), and the boxes subjects were previously assigned to reversed.  Reinforcement (food pellet for correct lever response) lasted for a 15 - minute time segment for 28 days.  Extinction followed, and no food pellets were given in either condition while subjects were tested for a two-minute time period in each condition.  Results provided evidence that subjects could discriminate between complex auditory stimuli.  There was no significance found for subject’s discrimination preference between musical conditions.  A significant context discrimination was found, as well as Box preference.  The implications of these findings are analyzed in the Discussion.  This study promotes further research into what effects auditory stimuli and context has on learning.

Back to top

Introduction

    Although many studies have provided research that rats can discriminate between general audio stimuli, their ability to discriminate between complex audio stimuli has been studied less frequently (Poli, 1991).  Research by Neuringer and Porter (1984) provided evidence that pigeons could discriminate between two complex auditory stimuli.  Because rats and pigeons tend to have similar learning capabilities it was expected that rats would also be able to discriminate between complex auditory stimuli.  It is necessary to study complex audio stimuli so that knowledge can be gained on the effect of complex audio stimuli on learning (Joseph, 1982).  The purpose of the present study was to investigate whether rats could discriminate between two complex stimuli and what affects each particular stimuli would have on learning.
     Research has been conducted to explore the question as to whether animals can distinguish between two different music forms.  In a study conducted by Neuringer and Porter (1984) pigeons were given two different pieces of classical music. Two pigeons were presented it the S+ condition (the stimuli being J.S. Bach’s Prelude in C Minor for Flute) and were reinforced with a grain after the desired behavior of pecking occurred.  The same two pigeons were then presented with the S- condition (P. Hindesmith’s Sonata), and were given no reinforcement.  Results provided evidence that the pigeons were able to discriminate between the two pieces of music after only thirty-three sessions.  This study provides evidence that pigeons can discriminate between 2 complex audio stimuli.  Because rat’s audio learning capabilities are similar to pigeons audio learning capabilities this study provides evidence that rats may be able to discriminate between two complex audio stimuli.
Research conducted by Okaichi and Okaichi, (2001) evaluated whether rats could discriminate between complex auditory stimuli.  Subjects were put into a Skinner Box, and half of the rats were exposed to the song “Yesterday” by Beatles, the S+ condition (reinforced by a food pellet), and white noise, the S- condition (no reinforcement).  The second half of the rats were also exposed to the song “Yesterday” by the Beatles but it was the S- condition (no reinforcement), and white noise the S+ condition (reinforced by a food pellet).  After the above phase was learned by both groups, the S- condition was changed, in both conditions, to a piece of music by Mozart. The S+ condition stayed the same.  Although results showed that subjects in both groups responded correctly to both stimuli, correct response in the S+ white noise condition was learned faster than in the S+ “Yesterday” condition.  This research provides evidence that although rats can discriminate between music and noise, it is more difficult for them to discriminate between melodies.
     The purpose of this study was to answer the question as to whether rats would have a greater response rate of target behavior when stimulated by a piece of music in the allegro tempo in the major mode than when they were stimulated by a piece of music in adagio tempo in the minor mode.  It was predicted that a distinction would be made between whether (1) subjects would be able to discriminate between the two complex stimuli, and whether (2) subjects would have a greater response rate for target behavior when stimulated by the allegro tempo piece of music in major mode.

Back to top

Method

Subjects:
    Eight Male rats were used in this experiment.  The rats were 80% free feeding weight and weigh from 280-230 grams.  These Sprague-Dawley rats came from Harlen Breeders in Indianapolis, IN, in May of 2001.  The rats were housed in the Psychology department at Saint Anselm College in Manchester, NH, in a 12-hour light/dark cycle.  The rats have been previously trained in a drug discrimination study and are pre-trained to left lever press in an operant chamber.  The rats were used in accord with Saint Anselm’s Animal Use Committee.

Apparatus:
     Subjects were placed in eight stainless-steel operant chambers (Med-Associates, Georgia VT, model ENV-001) measuring L 28 x W 21 x H 21 cm.  Each chamber contained one lever located 2 cm to the left and one lever to the right of the centrally-located food magazine (which delivered 45 mg food pellets, PJ Noyes, Lancaster NH) and 7 cm above the grid floor.  The chambers were spaced two to three feet apart about the perimeter of the sound- and light-attenuated experimental conditioning room measuring L 16.5 x W 9 feet.  Low level (approximately 15 watts) overhead incandescent lights acted as the light source during and at the start of the session. Experimental events were programmed via Med-PC Software (Version 2.08) and by a DIG interface (Med-Associates, Georgia, VT) to a 386 PC.

Procedure:
    After the 6-day pre-training time period subjects were introduced to two different music conditions, five days a week, Monday through Friday for a fifteen minute time period.  Subjects were trained for a 28-day time period on variable-interval schedule.  Subjects were randomly assigned by the flip of a coin to one of the two conditions, Condition A (Samuel Barber’s Adagio for Strings), or Condition B (Rachmaninoff’s Rach Second Symphony Allegro).  No subject participated in the same condition more than two days in a row.
     Music Discrimination Test
    Subjects were then tested for a 4-day time period.  On the first and second testing day subject 1, subject 2, subject 7 and subject 8 were tested simultaneously directly followed by subject 3, subject 4, subject 5 and subject 6.  Subject 1 and subject 7 were assigned to Box A, and Subject 2 and Subject 8 were assigned to Box B.  They were then tested for correct and incorrect responses under Condition A during a two-minute extinction time period, where no reinforcement (food pellets) was given.  The above subjects were then put back on the training schedule described in the training session paragraph.  Subject 3 and subject 5 were placed in Box A, and subject 4 and subject 6 were placed in Box B, they were then tested for correct and incorrect responses under Condition B during a two-minute extinction time period, and no reinforcement (food pellets) were given.  The above subjects were then put back on the training schedule described in the training session paragraph.
On the second testing day testing conditions were reversed (subject 1 and    subject 7 were placed in Box B, and subject 2 and subject 8 were placed in Box A.  Subjects were tested under Condition B.  Subject 3 and subject 5 were assigned to Box B, and subject 4 and subject 6 were assigned to Box A and tested under Condition A.  Correct and incorrect responses were measured in a two-minute extinction time-period.  All subjects were then put back on the training schedule described in the training session paragraph.
     Context Discrimination Test
    On the third and fourth testing day subject 1, subject 2, subject 3 and subject 4 were simultaneously tested together, directly followed by subject 5, subject 6, subject 7, and subject 8 who were simultaneously tested together.  Subject 1 and subject 3 were assigned to Box A and subject 2 and subject 4 were assigned to Box B.  All subjects were tested for a 2-minute extinction time period during which no music was played.  Subject 5 and subject 6 were then assigned to Box A and subject 7 and subject 8 were assigned to Box B.  All subjects were tested for a two-minute extinction time period during which no music was played.
    On the fourth day of testing subject 1 and subject 3 were assigned to Box B and subject 2 and subject 4 were assigned to Box A and were tested for a two-minute extinction time period during which no music was played.  Subject 5 and subject 7 were then assigned to Box B and subject 6 and subject 8 were assigned to Box A, and were tested for a two-minute extinction time period during which no music was played.

Back to top
Results

     Music Discrimination Test
    For each of the eight subjects discrimination ratios were calculated for testing under Condition A (piece if adagio music) and Condition B (piece of allegro music).  The number of correct responses (CR) divided by the sum of the number of correct responses and the number of incorrect responses (IR) during the two-minute testing session is how the discrimination ratios were calculated.  The discrimination ratios were calculated using the following formula:  CR/(CR+IR).
     A discrimination ratio is significant at or above .80.  5 of the subjects had significant ratio discriminations in Condition A.  The 3 remaining subjects all scored above chance (.50) in Condition A.  3 of the subjects had significant ratio discriminations in Condition B.  The remaining 5 subjects all scored above chance in Condition B (see table 1).  A paired samples t-test was then conducted to see if there was a significant difference between correct response percentages in Condition A and in Condition B.  There were no significant differences.
     Context Discrimination Test
    Subjects were then tested to see if there was context discrimination between the two operant chambers, Box A (correct response was left lever pressing) and Box B (correct response was right lever pressing).  The same method of using a discrimination ratio was used to calculate whether subjects made a context discrimination.
    Five of the subjects made a significant context discrimination in Box A.  1 of the remaining 3 subjects scored above chance in making a context discrimination in Box A.  7 subjects made a significant context discrimination in Box B.  The remaining subject scored above chance in a context discrimination ratio in Box B (see table 2).  A paired samples t-test was conducted to see if subjects made a context discrimination between Box A and Box B.  There was a significant difference in subject’s discrimination between Box A and Box B [t(7) = 2.095, p= .07].  The means revealed that there was a higher rate of correct response in Box A (M = 97.25, SE = 2.75) then in Box B (M = 72.87, SE = 11.50) (see table 2).


Back to top

Discussion

    The purpose of this study was to examine whether eight previously trained rats would have a greater response of target behavior when stimulated by a piece of music in the allegro tempo in the major mode than when they were stimulated by a piece of music in adagio tempo in the minor mode.  It was predicted that a distinction would be made between whether (1) subjects would be able to discriminate between the two complex stimuli, and whether (2) subjects would have a greater response rate for target behavior when stimulated by the allegro tempo piece of music in major mode.
     Complex Auditory Discrimination
    Each of the 8 subjects could discriminate between each of the musical conditions.  No subject had more incorrect responses then correct responses in either musical condition.  Under Condition A, the adagio musical condition, 5 subjects responded with a discrimination ratio of over 80%.  Under Condition B, the allegro musical condition, 4 subjects responded with a discrimination ratio of over 80%.  This finding suggests that these subjects had learned to discriminate between complex auditory stimuli.  The remaining subjects in both Condition A and Condition B had a discrimination ratio above chance (50%).  This suggests that these subjects were learning to discriminate between the two complex auditory stimuli, and if trained for a longer period of time would have made a greater discrimination between the two complex auditory stimuli.  These findings are consistent with Joesph (1982) research that rats can discriminate between complex auditory stimuli.
    There was no significance found when a paired samples t-test was conducted comparing discrimination between Condition A and Condition B.  This suggest that although subjects could discriminate between the tempo and mode of the particular piece of music to elicit correct response, neither the adagio or the allegro pieces of music had a direct affect on learning.  Rescorla (2000) found that variations in stimuli would have different effects on the percentage of a subject to the target behavior.  Rescorla’s (2000) study provided research that subjects in an excitory stimuli condition would have a greater percentage of response rate than subjects in the inhibitory stimuli.  In the present study, both pieces of classical music were excitory stimuli and therefore elicited the same amount of response from the rats.  This could be why there was no significant difference between correct response in Condition A and Condition B.
    Research conducted by Okaichi, and Okaichi (2001) provided evidence that although rats could discriminate between complex auditory stimuli, it is more difficult for them to discriminate between two different complex melodies.  Both pieces of music played to subjects in the present study had complex melodies.  Although subjects could discriminate between the complex auditory stimuli to elicit significant correct responses, the auditory stimuli in both conditions might have been to complex for subjects to have a preference for either of the conditions over the remaining condition.   This could be an indicator as to why subjects showed no significant preference between music conditions.
     Context Discrimination
    A context discrimination was found between Box A and Box B.  Results provided evidence that 6 out of the 8 subjects discriminated between both Box A and Box B over 80% of the time.   These findings are consistent with Thomas, McKelvie, and Mah (1985) study as to whether pigeons could learn contextual cues.  Both auditory and visual stimuli were used, and researchers found that pigeons could learn context cues in the presence of auditory cues.  Because research by Neuringer and Porter (1984) has provided evidence that pigeons and rats have similar learning capabilities, findings from the present study support Thomas, McKelvie, and Mah (1985) research that contextual cues as well as auditory stimuli play a role in discrimination.
    Subjects also had a significant rate of response between Box A and Box B.  When tested in Box A, subjects had a higher rate of correct response than when tested in Box B. However, this preference could have occurred because of the subject’s previous training history.  Subjects were predisposed to respond to left lever pressing (Box A).  However, this suggests that when the music (both Condition A and Condition B) was playing, subjects previously learned behavior (predisposal to left lever pressing) was extinguished, but when there was no music playing, subjects regressed to the behavior that they were predisposed to.  This suggests that the music did have some effect on learning.  What affect the music had on learning can not be determined by the findings in the present study, and further research should be conducted to explore this question.
     Conclusion
     The findings in the present study provide evidence that rats can discriminate between complex auditory stimuli.  However, there was no significant correct response rate between Condition A and Condition B.  This may be because the pieces of music in both conditions were excitory stimuli (Rescorla’s 2000).  Studies conducted in the future should use an excitory stimuli and a inhibitory stimuli to test whether subjects would have a significant between condition response.
      This study provided evidence that music does have some effect on rats learning.  When the rats were played music they did not regress to their previously learned behavior, but when there was no music played, rats did regress to their previously learned behavior.  It is unknown why this occurred and what affect the music had on learning.  This finding suggests that further studies should be conducted that explore what affects music has on previously learned behavior.
     This study not only provided evidence that rats could discriminate between complex auditory stimuli, but that rats could make contextual discriminations as well.  The findings in this study also create many questions that have yet to be researched (i.e. what effects music has on previously learned behavior).  It is necessary to conduct further studies exploring these questions.

Back to top

References

Joseph, C. (1982).  Effect of music on the behavioral organization of albino rats using the operant conditioning technique.  Indian Journal of Applied Psychology, 19 (2), 77-84.

Neuringer, A. & Porter, D. (1984).  Music discrimination by pigeons.  Journal of Experimental Psychology: Animal Behavior Process, 10, 138-148.

Okaichi, Y. & Okaichi, H. (2001).  Music discrimination by rats.  Japanese Journal of Animal Psychology, 51, 29-34.

Poli, M. (1991).  Discrimination of musical stimuli by rats (Rattus norvegicus).  International Journal of Comparative Psychology, 5 (1), 7-18.

Rescorla, R. (2000).  Associative Changes in Excitors and Inhibitors Differ When They are conditioned in Compound.  Journal of Experimental Psychology: Animal Behavior Process, 26 (4), 428-438.

Thomas, D., McKelvie, A., & Mah, W.  (1985).  Context as a conditional cue in operant discrimination reversal learning.  Journal of Experimental Psychology: Animal Behavior Processes, 11 317-330.

Back to top






Appendix

Table 1:  Music Discrimination Ratios

                          Adagio                                                               Allegro

Correct         Incorrect         %                               Correct         Incorrect       %

S1             37                  10               79                                         28                      2             93
S3             42                    9               82                                           6                      2             75
S5             33                    5               87                                         13                      2             87
S7               5                    5               50                                           5                      0           100

S2                2                   0              100                                          2                      1             67
S4                2                   1                67                                         10                     3             77
S6              19                   2                90                                         12                     8             60
S8                5                   0              100                                           5                     3             63

Note:  The adagio condition is referred to as Condition A, and the allegro condition is
            referred to as Condition B throughout the paper.
 
 

Table 2:  Context Discrimination Ratios

                       Box A                                                                     Box B

Correct         Incorrect         %                               Correct         Incorrect       %

S1             12                     0              100                                        22                      3             88
S3             29                     0              100                                          2                      0            100
S5               9                     0              100                                        26                      2              93
S7               6                     0              100                                          8                      1              89

S2                5                    0              100                                          2                      0             100
S4              36                  10                78                                          4                      2               67
S6              10                    0              100                                          1                      6               14
S8                1                     0             100                                          1                      2               33

Note:  In Box A left lever pressing is the target behavior, and in Box B right lever pressing is the target
           Behavior.

Back to top