Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
EXTENDED PRIMARY AND HIGHER ORDER CONDITIONING OF EARTHWORMS by THOKtAS ALVIN FIELDS, B.A, A THESIS IN PSYCHOLOGY Submitted to the Graduate Faculty of Texas Tech University in Partial FulfiTi.,.'--nt of the Requirements for the Degree of MASTER OF ARTS Approved August, 1970 ^ m AeH-3«l?' ^ ACKNOWLEDGMENTS I am deeply indebted to Dr. Sam L. Campbell for his direction of this thesis. Without his patience, critical direction, and personal involvement, this research might never have been completed. I would also like to express my graditude to Mr. Robert 0. Haynes for his help in constructing and maintaining my apparatus. Mr. Haynes was also invaluable in the computer compilation of the data. IT CONTENTS Page ACKNOWLEDGEMENTS ii LIST OF TABLES iv LIST OF FIGURES v I. INTRODUCTION 1 Instrumental (Operant) Learning Studies . 1 Classical (Pavlovian) Conditioning Studies 3 Purpose II. 7 METHOD 9 Subjects 9 Apparatus 9 Procedure and Experimental Design . . . . III. RESULTS AND DISCUSSION 14 19 First-Order Conditioning (Phase One). . . 19 Second-Order Conditioning (Phase Two) . . 22 and Re-Conditioning (Phase Three). . . 22 Second-Order Extinction (Phase Four). . . 26 First-Order Extinction (Phase Five) . . . lY. CONCLUSIONS 26 30 REFERENCES 32 m LIST OF TABLES Table Page I. Stimulus Configurations Per Group and Number of Trials Per Day in Each Phase IV 16 LIST OF FIGURES Figure 1. 2. 3. 4. 5. 6. 7. 8. Page Experimental chamber with access door open, showing conditioning trough. Portable control panel is to right and connector plug to Automatic Programming Instrumentation to lower left 10 Close-up of conditioning trough on grid platform showing electrodes and leads at either end, resting on millimeter grid paper and supported by fiberboard base with speaker mounted below 12 Number of conditioned responses per group per day during firstorder conditioning 20 Average latency of conditioned response per group per day during first-order conditioning 21 Number of conditioned responses per group per day for secondorder conditioning 23 Number of second-order conditioned responses per group per day on test trials during second-order conditioning 25 Number of second-order conditioned responses per group per day during second-order extinction 27 Number of first-order conditioned responses per group per day during first-order extinction 29 CHAPTER 1 INTRODUCTION The earliest reported behavioral studies of earthworms was executed in 1889 by Hesse (Jaeobson, 1963). Hesse reported that bodily contractions, interpreted as withdrawal responses, occurred with decreases in illumination. Smith (1902),in a replication and extension of Parker and Arkin's research (1901),investigated the effects of temperature, odors, light, and tactual stimulation on contractual behaviors of worms. Shortly thereafter, investigators began utilizing instru- mental operations in the investigation of worm behavior (Yerkes, 1912; Heck, 1920; Robinson, 1953; and Datta, 1963). Instrumental (Operant) Learning Studies In 1912, Yerkes demonstrated that the earthworm could discriminate between an arm of a T-maze which lead to an earth-filled goal box and an arm which lead to electro-shock. Heck (1920) replicated Yerkes' study but used a larger number of subjects and three varieties of earthworms. He found that the worm required 100-200 training trials to discriminate the correct arm of the T-maze, and an additional 50-70 trials to learn a reversal of the original discrimination. Robinson (1953) questioned Yerkes' conclusion that stimuli other than those normally associated with a maze were required just prior to encountering shock, to facilitate learning. 1 Using a T-maze, Robinson reported that there were two distinct phases in the learning of an operant task by earthworms. The first phase occurred about trial 50 when "running time" significantly increased. Robinson described the worm at that phase as reacting negatively to all portions of the maze. By "negative", he could be referring, quantitatively, only to the temporal changes in response properties. But Robinson concluded that the first phase decrements were attributable to classically conditioned withdrawal responses which hindered the acquisition of instrumental (operant) responses. He also suggested that when training was carried beyond the first phase, no stimuli other than those ordinarily in the maze were required for learning. Datta (1963) taught a position habit in a T-maze by presenting shock at the terminus of one arm and the home container at the terminus of the other arm. Her conclusion: a one-minute intertrial interval re- sulted in the same degree of learning as a five-minute intertrial interval. Longer intertrial intervals, such as 25 minutes, resulted in no learning. Also progressive improvement was obtained after several training trials. Other studies utilizing instrumental operations to investigate earthworm behavior include Swartz(1930), Wherry and Sanders (1941), Bharueha-Reid (1956), Krivanek (1956), and Arbit (1960). There have been several studies which proposed to show that the behavior described in the above experiments was not attributable to learning but to some other behavioral process such as pseudoconditioning or sensitization. Fisehel (1933) was unable to train worms to use tactual cues which were provided in a Y-maze. However, the assumption was made that the tactual cues would be the only relevant cues in the situation and no attempt was made to control other relevant stimuli. Fraser (1958) in a much more comprehensive investigation questioned methodological aspects of prior studies after failing, himself, to obtain learning in a T-maze. He proposed that the failure by earlier ex- perimenters to specify pretraining turning preferences and then to train the worms appropriately to the non-preferred side, substantially weakened their claims that learning had been obtained. Fraser also suggested that the criteria of learning utilized were so low (weak or permissive) as to admit factors other than learning to be more relevant to obtained changes in behavior. Kirk and Thompson (1967) obtained no evidence of learning as measured by "crawl-time" in a straight alley, and simply questioned whether instrumental learning by earthworms was possible. Classical (Pavlovian) Conditioning Studies With the exception of three studies in the 1930's (Copeland, 1930; Copeland and Brown, 1934; and Raabe, 1939), all research involving Pavlovian conditioning of earthworms has developed since 1959. Perhaps this is attributable, as suggested by Bitterman (1960), to the difficulty of recording classical responses. Certainly, the tender membranes of the worm cannot be operated upon with the crude meehaneial devices so often required when working with a new species (i. e., one for which standard laboratory instrumentation is not available), nor can it be expected at this time, that institutions regulating psychological research would invest in researches about an organism so diverse from man. Copeland (1930) was the first to use classical conditioning techniques as a basic paradigm in earthworm research. The worm was housed in a water-imnersed tube and obtained food from the water by characteristic, bodily-involved, ingestive movements. Copeland preceded the presentations of food by an increase in illumination. Within five trials, food- gathering responses occurred at the onset of the light. Copeland and Brown (1934), using the same general experimental technique, substituted a tactual cue (touch to the anterior end of the worm) for the light cue. Again, reliable conditioned responses were obtained after a few trials. Raabe (1939) found that shock, tactual cues, and vibrations could serve as unconditioned stimuli. By preceding any of these unconditioned stimuli with an increase or decrease in illumination, a conditioned withdrawal response ("retraction" would be the better term to use, since we refer to classical reflexes, and "withdrawal" might be confused with instrumental behaviors) could be established in 10-20 trials. Raabe also argued that contiguity was critical in the establishment of the conditioned response. He found the most effective conditioned stimulus-unconditioned stimulus interval to be one-half to one second. Intervals as long as three seconds were less effective; intervals of four and zero seconds were ineffectual. Finally, he found that the onset of light was more successful as a conditioned stimulus than was the offset of light. Ratner and Miller (1959) published what Jaeobson (1963) called the first study of true classical conditioning in the earthworm. Ratner and Miller conditioned a one-eighth to one-quarter inch withdrawal response to vibration with light serving as the unconditioned stimulus (UCS). They used a 50 second intertrial interval. The vibratory conditioned stimulus (CS) was on for six seconds with the illuminative UCS on for the terminal two seconds. One hundred conditioning trials preceded extinction. The earthworms showed a significant increase in the conditioned response (CR) to the CS during training and a significant decrease of CR to the CS during extinction. Three control groups were used. The first control received 100 trials of six seconds of vibration without an UCS. The second control received 100 trials of four seconds of observation without the CS or UCS. The third control received 105 trials consisting of 70 light and 35 vibration trials. The two second UCS was given in blocks of ten trials followed by a block of five six-second CS trials. Bitterman (1960) mentioned pairing a neutral stimulus with shock which produced a conditioned withdrawal reaction in the worm. He at- tempted to measure the response on a kymograph with little success. He did note that the response was clearly visible to the naked eye. Herz, Wyers and Peeke in 1963 evaluated the effects of partial reinforcement on a classically conditioned response. Using a vibratory CS, one group received the UCS of light on 100 percent of the acquisition trials. The second group received the UCS on a random 50 percent of the acquisition trials. There was no difference between groups during acquisition but the partial reinforcement group showed increased resistance to extinction. Peeke, Herz and Wyers (1964) used light as an UCS and vibration as a CS to investigate the effects of partial reinforcement, and amounts of acquisition training on resistance to extinction. The CR was initially described as a one-half to one inch withdrawal of the anterior segments, or stoppage and/or rearing of the anterior segments of the body. But since the rearing response occurred more reliably, it was subsequently used as the critical response. were utilized: (1) Six conditioning and two control groups 100 percent reinforcement with 50 acquisition trials, (2) 100 percent reinforcement with 100 acquisition trials, (3) percent reinforcement withl50 acquisition trials, (4) 100 50 percent rein- forcement with 50 acquisition trials, (5) 50 percent reinforcement with 100 acquisition trials, (6) 50 percent reinforcement with 150 acquisi- tion trials; the control groups were a sensitization control where the UCS was omitted on every trial and the pseudoconditioning control where the CS was presented alone for the first ten trials and then the UCS was presented alone for the next five trials. All groups were trained over three days. The results indicated no differences among groups with respect to the ratio of CR's to total number of acquisition trials. But the partial reinforcement group showed increased resistance to extinction when compared to the continuous reinforcement group. Wyers, Peeke and Herz (1964) replicated the study using light as the UCS and vibration as the CS. Again, they investigated the effects of partial reinforcement. The results were the same: the partial re- inforcement group showed increased resistance to extinction. Morgan (1965) and Morgan, Ratner and Denny (1965) attempted to measure a Galvanic Skin Response (GSR) to onset of illumination. The worm was stitched to a damp sponge with the GSR electrodes mounted through opposite ends of the worm. Five groups of subjects were used. The first group consisted of dead controls used to adjust the apparatus The other groups received 480, 180, 80, or zero footeandles of illumination. The authors obtained reliable indications of Galvanic Skin Response. As the intensity of the stimulus increased, the intensity of the response increased. The 480 footcandle group made the largest responses and showed the slowest adaptation to the stimuli. However, all subjects adapted to the stimuli in three to five trials. Evans (1966) has argued that evidence for the classical conditioning of earthworms is, at best, tenuous, and that adequate controls for pseudoconditioning and sensitization have not been utilized. He noted that earthworms respond to many stimuli with withdrawal responses, and that response magnitude varies directly with stimulus magnitude. His assertation accompanied a study in which he was unable to find evidence of learning after 150 conditioning trials. Purpose A first objective of the present study is to evaluate Evans' assertation regarding the lack of evidence for classical (Pavlovian) conditioning of earthworms. I intend an extensive study utilizing far more than 150 conditioning trials, utilizing adequate controls for both 8 sensitization and pseudoconditioning, and, further, evaluating any conditioning results with subsequent extinction operations. A second objective of this study is to inquire whether higherorder classical conditioning can be obtained with earthworms - no evidence of an attempt to do so appears in the literature. Pavlov (1960) described a conditioning program whereby an original CS was used as a substitution for an UCS in the conditioning of a second CS. A dog was conditioned to salivate to the sound of a metronome. A black square, which elicited no salivation, was held in front of the dog and the metronome sounded. After ten trials, the dog salivated to the presentation of the black square. Pavlov referred to this as second order classical conditioning. CHAPTER II METHOD Subjects Twenty-four earthworms (Annelida-Lumbricus terrestris) varying In length from four to five inches were obtained from a local bait shop. They were housed individually in one pint plastic containers filled with a moist mixture of coffee grounds, grass, moss, and earth. The containers were stored on semi-dark shelves at room temperatures ranging from 75 to 78 degrees F. Apparatus Exploratory studies indicated that S^s could not survive frequent handling. Accordingly, apparatus was designed so that each of six Ss could be placed in its own experimental chamber and remain (without handling) for a series of conditioning trials over a daily training session. To initiate appropriate combinations of stimuli (electroshock, light, and/or vibration) and to record S^'s response, the experimenter carried a programming panel which could be plugged into each experimental chamber successively (Fig. 1). Through the programming panel, behavior was recorded and the selection, timing, and sequencing of stimuli were automatically controlled by instrumentation housed in an adjoining, sound-proofed room. Moving from chamber to chamber with the control panel permitted maximal use of recording and control facilities while minimizing handling of S^s and allowing for adequate intertrial intervals. 10 0) I— I •M E Xi C $ - =3 O 0) cn S- 11 Conditioning trough: The conditioning trough was the bottom half of a clear plastic toothbrush case to which electrodes were attached at opposite ends (Fig. 2). It rested on a grid-covered fiberboard platform, to the bottom of which was bolted a four by six inch oval speaker. The platform, in turn, rested on sponge-rubber supports to Isolate it from extraneous vibratory stimuli. Prior to each day's training session, Ss were rinsed in room temperature tap water and the troughs moistened. Experimental Chamber: The experimental chambers were twelve inches wide by six inches deep by twenty-eight inches high, and were painted flat black. Each chamber housed a trough unit with its appropriate sources for shock and vibratory stimuli and had a 200-watt lamp mounted 18 inches above the trough platform for photic stimulation. A hinged door provided access to the interior and contained an observation port In its face. The observation port was covered with X-ray film and constant interior illumination was provided by a one-watt red lamp mounted nine inches above the trough. Hess (1924) found this photic stimulation to be subliminal for earthworms. Stimulus Properties: During first-order conditioning of all groups, the unconditioned stimulus was a .16 milliampere electro-shock of one second duration. It was preceded by a 200-watt light for two seconds, and both light and shock were present during the third second of each trial. 12 (/) -o «0 "O <u s. 10 •~ o -O.Q c slO 0) v\ 9) (f-o > * o S.XI 4-> o -a <u 0) •!-» r— (U o O) o . c CL .r" =) 5 to o JC="t3 (O c: (0 SO 0) «»- CL (O +> (O CL ^— g D--0 •o •r^ .r- s- s. C7> cn $- 0> c +J o f <D E r— cn 3 r~ • 2 o r— O -r- 0) s. E - Q 4J C T3 C7) O 0) C +J .r— cn 3c c o o .r— • M E 4-> V> •f— a s• o s> 0) c * .^ o <o <u o-o c o. «»- Q) cn o S. JC C •f— <U 4-* = J ^ .f— t +J S 0) . r ~ tn 0) 0) O (/) 4-> ra O (Q ^ OL CNJ 0) s3 C7> 13 During second-order conditioning of the experimental S^s, a 60 cycle vibratory stimulus was presented for two seconds and then overlapped the light for an additional second. Various combinations and orders of presentation of these stimuli were used with the several control groups described below. Excepting extinction trials and the occa- sional test trials interspersed during conditioning, every trial consisted of the three-second presentation of a stimulus overlapping a stimulus of different modality during the last second. In all instances, whether conditioning, test or extinction trials, 90 seconds elapsed between trials for a given S^. The unconditioned response to shock is characteristically a violent contraction and coiling. In this study if S^s were elongated at the ini- tiation of a trial, contraction of both the anterior and posterior ends was considered evidence of an adequate conditioned response. If the S were already partially coiled, then additional constriction of the coil and withdrawal of the head into the coil was considered a conditioned response. Response Definition and Recording Procedure: Because of difficulties previously encountered in attempts to automatically record respondent reflexes of earthworms (Bitterman, 1960), Ss' responses were observed by the experimenter and were recorded by both an Ester!ine-Angus event marker and a Foringer print-out counter when the experimenter depressed the "response" switch on his portable control panel. In the analysis of records, a response was judged to be a conditioned response 14 only If It occurred during the first second of the initial stimulus the stimulus to be conditioned, whether in first-order or second order conditioning. The simultaneous and automatic recording of onset of the various stimuli on separate event-marking channels indicated the latency of each response. As a check on the reliability of the experimenter's discrimination of conditioned responses, other observers were given a verbal description of the experimental procedure and conditioned response characteristics. They were then asked to make independent judgements as to whether a conditioned response (CR) occurred on given trials. A Pearsonian correlation of .91 was obtained between judgements of the experimenter and independent observers. Procedure and Experimental Design S^s were run in four waves per day, six S^s per wave (one S^ for each experimental chamber). The v/ave to which S^s were assigned was randomly determined, but remained the same for given S^s throughout training so as to control for elapsed time between sessions. The 24 S^s were equally but randomly assigned to one of four groups prior to the first-order conditioning to which all Ss were subjected. Assignment to four different groups was intended to provide information regarding variability of conditioning trends among groups of six subjects during first-order conditioning. Were the results sufficiently variable, they might preclude expectations of differences attributable to various treatments provided during succeeding phases of training. 15 During first-order conditioning, all S^s were given 20 trials per day with light as the to-be conditioned stimulus and shock as the unconditioned stimulus. Following the 16 days of conditioning which provided a total of 320 trials, and despite the yery similar and overlapping performances during days 12-16, S^s were again equally but randomly as- signed to four groups and given an additional session of first-order conditioning to evaluate the comparability of the basal performance with which various groups entered the second phase of training. The four groups of subjects were trained in five phases, with stimulus operations varying among groups and from phase to phase. These variations were intended to provide controls for sensitization and pseudoconditioning, and tests of conditioned response strength by subsequent second- and first-order extinction. Table 1 indicates the stimulus operations for various training phases of the four groups. The groups are designated Experimental (Group I ) , Light Sensitization Control (Group II), Random Control (Group III), and Shock Control (Group IV). The phases can be generally described as: Phase One, first order conditioning (340 trials over 17 days), identical for all subjects as described above; Phase Two, second-order conditioning or control for same (200 trials plus 60 test trials); Phase Three, first-order re-conditioning or conditioning to vibration as a conditioned, new firstorder, stimulus for Groups II and IV, respectively (80 trials over four days with no test trials interspersed); Phase Four, second-order 16 u o Ol > .£= C/) I 4-> •4-> x: cn x: cn o o O u. c o •r•M fO O CM c o •^• sz o c o 4-> 4-> (U .Q •r— <o a. > J:: •r— «o s- s- &- to cn O CM CM s3 4J x: cn o OJ u a> .o X) X) X) o o o CsJ CM CM o o x: to E 3 •ZL -o c «TJ CL 3 O &. o &- »—1 Ul -J CO «c 1— a> OV) c/> x: Ou x: o as E X) E o cvi c •r- o »T3 to CM >> «a o 4-> (O S0) Otn H-* c o o •r.i^ t— O I M- o c c - c o o o o tn S> to T— -r— 1— CO 03 4 J 03 nj x : i- s- cn-Q J3 •r— •!— T— I— > > 3 O E O ^ c o c c: o o cn fO •r- C I c: o O) Ul c o .r— s3 cn •r*-* o o x: to I +J x: cn o SX3 C •1- O > 'rI -M •4-> fC x : s- cnxj I— > o ^ •r+J o cn c: o x> o to > > I I I rO SXI •I- > •«I .^ m o $- O XJ O JD x : -rto > cn cn o I— f— CO to > i n l o t n LO ^ O -slCM CVJ c o to <t5 O. 3 O o cn o o x: to I x: cn o o o o C\J CM CM CM CO I x: cn u o sz <A I 4-> r~ ••-> +-> c x : cj o s- CO o o x: E O O x: (/) I 4J x: cn 17 extinction (Three seconds of vibration for 80 trials over four days). Identical for all groups; Phase Five, first-order extinction (Three seconds of light for 200 trials over 10 days). Identical for all groups excepting Group II which served as a control for extinction of other Ss and continued to receive light and shock conditioning during Phase Five. Rationales for the various stimulus operations utilized among the groups and training phases can best be specified in connection with the Results of this report, but a detailed description of operations used in Phase Two and Phase Three conditioning of the control groups (II, III, and IV) follows. It should be understood that when reference is made to two stimuli which are hyphen-connected, the first stimulus was of three seconds duration and overlapped the second stimulus by one second. Light Sensitization Control (Group II): During Phase Two these Ss received 20 light-vibration pairings and four vibration (test) trials per day. The order of stimulus pairings was intended to determine if light sensitization would obtain (i.e., if light-vibration pairings could produce a second-order conditioned response during vibration test trials) or maintain a first-order conditioned response to lights. Following 15 days of Phase Two conditioning, these Ss were retrained during Phase Three with first-order conditioning procedures. Random Control (Group III): ^s received random combinations of all stimuli with the exceptions of vibration-light and vibration-shock to evaluate whether other combinations of stimuli might produce or sensitize 18 a second-order conditioned response. They were given five trials each of shock-light, shock-vibration, light-vibration, and light-shock pairings, randomly dispersed among four test trials of vibration only. These stimulus operations contained through 15 days of second-order conditioning (Phase Two), and, as a control, through four additional days of Phase Three conditioning. Shock Sensitization Control (Group IV): During Phase Two these S^s received 20 shock-vibration pairings randomly interspersed among four test trials per day. These stimulus operations were used to evaluate whether shock-vibration pairings would sensitize (produce) a conditioned response during the test trials. Following 15 days of Phase Two conditioning, the Phase Three conditioning comprised four days of first-order vibration-shock pairings to determine whether vibration could serve as an adequate conditioned stimulus. CHAPTER III RESULTS AND DISCUSSION First-Order Conditioning (Phase One) The reliable results of first-order conditioning is indicated in Fig.'s 3 and 4. Total number of conditioned responses per day for each group is negatively accelerated and approaches asymptote at about 115 CRs per daily training session. There is little variability but considerable overlap among groups, and especially during the last nine sessions when asymptotic performances produce more than 90 percent CRs per total trials. Krusal-Wallis tests do not provide sufficient evidence of differences among groups during the first or last four conditioning sessions (p is far greater than .05 for H=1.312 and H=1.022). But a Mann-Whitney U test (Siegel, 1956) indicates a significant difference between number of CRs during the first versus the last four days of conditioning (z=2.32, p .05) Randomization of all ^s into four "new" groups after the sixteenth daily session (see B, Fig. 3) resulted in almost identical total performance among groups during the seventeenth conditioning session. Finally, latency of CR data as graphed in Fig. 4 supposts the above analysis. 19 I " ( a o 4-> •?— f t a o sc> 01 o I c/> e. k it * u X cn -C5 ^1 OL O. 3 O s- cn s. ex. V) sz O a. CO <y s-o c o c o a T" o n—r "T" o 1—r—I—r—I" o o o o 00 CM 0) X3 e 3 S3SN0dS3a G3NOIiiaNO0 y39WnN a> i3 cn 21 V. <u -o so I CO X.f~. Ci- C7> c: I g H W 0M 3 Q- .. > ; fX 3 o -so i. -1 s.. c» _.to CL ~w (U to cn c o cn <u $~ CO •o -00 _to in ! - < • -lO -M ^ -a a» a o c: o u o >y (J c • a cn -«-> c . 03 T - r— ^ O Qi f- cn-M • • • • Q P *^ 05 •<— $- -a <u r: > o (SaN003S Nl) 3SH0d33a a3NOIJLiaN00 JO A0N31V-1 3 LL_ 22 Second-Order Conditioning (Phase Two) and Re-Conditioning (Phase Three) Re^P<^nses to the first stimulus on conditioning trials: The experimental S^s (Group I) show little evidence of second order conditioning by their response to vibration on the vibration-light trials during Phase Two and Phase Three (Fig. 5). But the criteria of a CR required that it have a low latency (occur during the first second of vibration) and may have been unduly severe. The critical evidence for or against second-order conditioning should appear on the interspersed three second test trials described below. The light sensitization control (Group II) demonstrates typical extinction of a first-order conditioned response once light-vibration pairings were substituted for light-shock pairings. Total CRs decline from about 100 to 10 over 15 days of Phase Two conditioning. But the original first-order CR is rapidly re-established when lightshock pairings are re-introduced during four days of Phase Three conditioning. The random control (Group III) maintained a relatively high and consistent number of responses to the initial stimulus under all four stimulus permutations (five pairings each per daily session). 23 0"> o •r^-> »«— I3> C -o c o c '"5 c o <J %V. X O <f4 1 -o o u CO 0) CO %~ o >) «? -o i- C3- o. 3 O SCn J<U CiC/) a; a to co CL i/> c; s- "TO Q) c o o ^- o i- x> E 3 S3SN0cJS3y 03NO!liaNO0 MSaWON ir> 3 cn 24 Since the shock sensitization control conditioning operations consisted of shock-vibration pairings (Group IV), responses to the initial stimulus would be unconditioned responses and are not graphed in Fig. 5 for Phase Two. But when it became apparent from test trial data (see Fig. 6) that sensitization had not occurred, these S^s were provided first-order conditioning pairings of vibration-shock to establish that vibration can serve as an adequate CS. The rapid conditioning of CRs to vibration during Phase Three training is evident at A, Fig. 5, where total number of CRs per session rises from zero to 75 over four daily sessions. This acquisition function is similar to that of original light-shock conditioning in this study and to the vibration-light conditioning data of Ratner and Denny (1959). Responses on Test Trials: The critical variable for evaluating second-order conditioning in this study is total number of CRs per group over the four test trials provided during second-order (Phase Two) conditioning, when vibration alone was presented for three seconds. All groups evidenced some degree of sensitization or pseudoconditioning during the first few days, but the experimental group alone continued to respond at a relatively high though variable level (see Fig. 6 ) . A Krusal-Wallis test for overall significance among groups provided a statistic which was slightly less than that 00 I' — V, ••••; I ^t sz o >> 03 ^ I H Wg K G. CL 3 O cn sCL »/> to - sz o CL i/> © . - - s> "O (U C C7> o c +J c .r- O -O •!E +J O 'r— o -o c u o iO $-. (U O I 01 CO c O sz o 0 -o cr._ u Qi o to s- cn X5 F 3 o3SNCdS3y G3ISiOIJ.iGMO0 daat^'ON VD 0) s3 cn fS3 26 required for significance at the .05 level (obtained H=7.27, while H=7.82 at p=.05). To locate the principle differences between groups, Mann-Whitney U tests were calculated and indicated that they were between Groups I and III and between Groups I and IV (U=60.5 for former, U=59.5 for latter, U=61 at p=.10), but barely significant at the p=.10 level of significance. Second-Order Extinction (Phase Fourj^ During Phase Four, vibratory stimulation alone was presented for three seconds, 20 trials per day over four days. Responses to vibratory stimulation by Group II (Light sensitization control) and Group III (random control) were negligible, as they had been during Phase Three training (see Fig. 7 ) . Group IV having been first-order conditioned by vibrationshock pairings during the prior Phase Three training showed a marked predictible decline in CRs when vibration alone was presented. However, consistent with evidence of second-order conditioning of the experimental group (I), there was an orderly decline from 17 to five total responses over the four extinction sessions. First-Order Extinction (Phase Five) The group II S^s which had been re-conditioned to first-order light-shock conditioning during Phase Three and were unaffected by exposure to vibration during Phase Four, were maintained on firstorder, light-shock conditioning as a control for the extinction 27 I •o c o o O) CO cn c: 3 >> -o icu CL O. 3 o scn sCL to GJ CO C o OL CO S- •o c O c o u (U ~r o T o 1^ o lO •t S3SN0dS3« a3N0lliaN00 U3Q\Hm -o sO I c o O •*— CJ + J cn C nr- O X a> sa; s. o <u EXJ 3 S2= O S3 cn 28 which all other groups underwent by the presentation of light alone. The overall differences among their performance (see Fig. 8) and that of other groups is statistically significant (KrusalWallis test yields H=14.04 with p<.01). Some rationale might be proposed for the most interesting absence of overlap among extinction curves of Groups I, III, and IV. While appropriate statistical tests would probably demonstrate the reliability of differences in these trends, statistical comparisons have not been made and any possibly relevant conclusions would be post hoc and not necessarily central to this thesis. 3 91 29 tc n o t> 3 1 HWS ^ 03 1 ii 44 •c? 1 CL CL 3 O J.. m s.. su -O 1'. Hi -o CL hi cn 0) cn - . CO \ / O CL to \ •X3 4 • -co W Si o CD / / ^ ^ SZ -tf> ^ / c o / / / f 1/ • « f h ir -5f / / / / -K) / -CJ / 1 •rvl ! 1 O •»-» o cn 1— i - +> .«- a «•- cO •r- s- li • -o o • ( c 1 1 O 00 1 ) o (0 j 1 1 1 o o ^ CM 1 1 O X O) a: u XI cu E-o 3 SZl o CO S3SN0dS3d aSMOILIONOO yi^QWON '«-' r-mr- CHAPTER IV CONCLUSIONS Considering such relevant factors as number of conditioning trials, controls for sensitization and pseudoconditioning, reconditioning and extinction controls, this study provides substantial evidence of first-order light-shock and vibrationshock conditioning of earthworms. While there have been implications that previously obtained data bearing on classical conditioning of annelida is somehow spurious (Evans, 1966), this experiment was not primarily designed to test the possibility of first-order conditioning of earthworms which was already felt acceptably demonstrated; rather it was intended to evaluate the possibility of second-order conditioning. Evidence of second-order conditioning obtained in this study was very nearly statistically reliable according to contemporary "standards" of the scientific community. But there is no statistic designed to evaluate the complex relationship among changing sequences of conditioning operations as employed in the present experiment. While considerable laboratory-control operations 30 JW- 31 were maintained in the collection of data reported, it is proposed that slight modifications in experimental design, especially sample size, might produce confident conclusions that earthworms can be second-order conditioned. REFERENCES Arbit, J. A failure to confirm "Latent learning in earthworms." Worm Runners' Digest, 1961, 3(2), 129-134. Bharucha-Re1d» R. Confirmation or refutation of latent learning in the earthworm? Worm Runners' Digest, 1961, 3(3), 179-183. Bitterman, M. E. Toward a comparative psychology of learning. American Psychologist, 1960, 15, 704-711. Copeland, M. An apparent conditioned response in Nereis virens. Journal of Comparative Psychology, 1930, 10, 339-354. Copeland, M. & Brown, F. A. Modification of behavior in Nereis virens. Biological Bulletin, 1934, 67, 356-364. Datta, Lois-Ellen. Some experiments on learning in the earthworm , Lumbricus terrestris. American Journal of Psychology, 1962, 75, 531-553. Evans, S. M. Non-associative behavioral modifications in the polychaete. Nereis divesicolon. Animal Behavior, 1966, 14(1), 107-119. Fisehel, W. Uber bev/ahrende und wirkende Gedachtnis-leistung. Biolos-ical Zbl. 1933, 53, 449-471. (In A. L. Jaeobson, Learning fn flatworms and annelids, Psychological Bulletin, 1963, 60, 74-94.) Fraser, C. H. T. Maze-learning in earthworms. Unpublished Master's Thesis, University of Aberdeen, 1958. (In A. L. Jaeobson, Learning in Flatworms and annelids. Psychological Bulletin, 1963, 60, 74-94.) Heck, L. Uber die bildung einer assoziation beim Regenwurm auf grund von dressurversuchen. Lotos, 1920, 68, 168-189. (In A. L, Jaeobson, Learning in flatworms and annelids. Psychological Bulletin, 1963, 60, 74-94.) o2 33 Herz, M. J., Wyers, E. S., & Peeke, H. V. S. Partial reinforcement of a classically conditioned response in earthworms. Newsletter for Research in Psychology. 1963, 5(3), 31-32. Hess, W. M. Reaction to light in the earthworm, Lumbricus terrestris. Jomrnal of Morphology and Physiology, 1924, O./J 010-OT'C. Hesse, R. Untersuchungen bei niederen thieren: A. Wiss. Zool. 1889, Learning in flatworms 1963, 60, 74-94.) uber die organe der lichtempfindungen V. Die augen der polychaten anneliden. 65, 446-516. (In A. L. Jaeobson, and annelida. Psychological Bulletin, Jaeobson, A. L. Learning in flatworms and annelida. Bulletin, 1963, 60, 74-94. Psychological Kirk, W. E., & Thompson, R. W. Effects of light, shock and goal box conditions on runway performance of the earthworm, Lumbricus terrestris. Psychological Record, 1967, 17(1), WW. Krivanek, J. 0. terrestris. Habit formation in the earthworm, Lumbricus Physiological Zoology, 1956, 29, 241-250. Morgan, R. F. A GSR technique for the response measurement of the earthworm. Papers of the Michigan Academy of Science, Arts & Letters, 1965, 5012)7337-342. Morgan, R. F., Ratner, S. C , & Denny, M. R. Response of earthworms to light as measured by GSR. Psychonomic Science, 1965, 3(1), 27-28. Parker, G. H., & Arkin, L. Directive influence of light on the earthworm, Allolabophora feetida. American Journal of Physiology, 1901, 5, 151-163. Pavlov, I. P. Conditioned reflexes: An investigation of the physiological activity of the cerebral cortex. Oxford: Oxford University Press, 1927~. Peeke, H. V., Herz, M. J., & Wyers, E. J. Amount of training, Intermittant reinforcement and resistance to extinction of the conditioned withdrawal effect in the earthworm (Lumbricus terrestris). Animal Behavior, 1965, 13(4), 566-570. 34 Raabe, S. Zur analyze der assoziationbildung bei Lumbricus variegatus. Z^. Vergl. Physiol., 1939, 26, 611-643. (In A. L. Jaeobson, Learning in flatworms and annelids. Psychological Bulletin, 1953, 60, 74-94.) Ratner, S. C , & Miller, K. R. Classical conditioning in earthworms, Lumbricus terrestris. Journal of Comarative and Physiological Psychology, 1959, 52, ! 02-105. Robinson, J. S. Stimulus substitution and response learning in the earthworm. Journal of Comparative and Physiological Psychology, 1953, 46, 262-266. Siegel, Sidney. Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill Book Company, 1956. Smith, Amelia C. The influence of temperature, odors, light and contact on the movements of the earthworm. American Journal of Physiology, 1902, 6(7), 459-486. Swartz, Ruth D. Modification of behavior in earthworms. Journal of Comparative Psychology, 1930, 10, 17-33. Wherry, R. J., & Sanders, J. M. Modifications of a tropism 1n Lumbricus terrestris. Transactions of the Illinois Academy of Science, 1941, 34, 237-238. Wyers, E. J., Peeke, H. V. S., & Herz, J. J. Partial reinforcement and resistance to extinction in the earthworm. Journal of Comparative and Physiological Psychology, 1964, blJVjy 113-116. Yerkes, R. M. The intelligence of earthworms. Journal of Animal Behavior, 1912, 2, 332-352.