CHAPTER 6

INKBLOT TASKS

by Hugh M. Lewis

A different set of projective inkblot tasks were employed which can be considered primarily projective in design, though a number of other tasks also have definite projective aspects to their elicitation. What is apparent is that projection may occur at different levels of elicitation and analysis.

A critical problem exists in the analysis and evaluation of these tasks, since such interpretation is formally held to be the purview of only qualified psychoanalysts. Since the theoretical ground for this type of analysis is for the most part outside of the main objective of this study, which is the culturally relevant symbolic content and patterns of response to such tasks, the system of analysis and interpretation here is different and independent of the standard form of analysis of these types of tasks. No attempt has been made to offer a psychoanalysis of the average or basic or "modal personality" of the Jetty people. Rather, these tasks were utilized as an interesting variant of the frame tasks which pose a similar type of problem for the subject in organization of the field in a structured and meaningful way.

The complexity of analysis of these samples renders a satisfactory presentation of their patterns beyond the scope of this work. Like the Luscher and other tasks already presented, these projective inkblots, when construed merely as devices for the elicitation of primarily projective "symbolic framing" responses, without the a priori ascription of universal or scientific significations to their interpretation, offer and interesting window for gathering various kinds of information which can then be organized and analyzed in a variety of alternate ways.

Scoring of the different tasks was standardized for the sake of consistency and comparison of frequency patterns across the samples. Locational scoring was done by dividing the space into an even 3x3 grid.

They were scored for whole and part responses, and an "part whole" response, as well as for minor detail responses. Scoring was also done for overall Form based upon a 1 to 5 rating scale, with five represent the clear and "true" use of form, and 3 about the average vague use of form, and 1 the lack of any clear resemblance between a prototypical object named and the location of the inkblot. Scoring was done for K-type responses in which a sense of 3-dimensional perspective, topographical overview or x-ray type responses seemed to emanate from a special "gestalt" recognized and thus named in the inkblot. "Ch" responses were included for the use of any implied shading in the image. M responses, which were infrequent, was for the implied use of action or animate movement of the image. Content scores included Human, Animal and Plant in whole, part or "object" categories, as well as for the use of space, abstracted "signs" or images, the use of pure shapes, such as "hearts or love" or triangles or squares or circles.

The resulting composite scores form a kind of frequency "psychogram" based upon the saliency patterns in the frequency scores across the different samples of informants. It is clear that in the analysis of these tasks there was considerable variation in the quality and strength of different subjects responses, but there was also a considerable amount of sharing at different levels. Color, form-color and color-form responses were scored only for the color version of the Harrower which also constituted the smallest sample, and such color responses were evident in only a few of this sample.

The five samples taken from the Jetty were mostly overlapping, based upon five different sets of inkblot pictures, two of which were of my own construction, one of which was a photocopied version of the Rorschach, lacking the color and fine detail, and another a similar version of the Harrower inkblot in both black and white and in color. The total sample size was about 135 sets of tasks. Given that the two "homemade" tasks had 15 pictures each, and the three others had 10 each, and that each one of these pictures generate several spreadsheets of scores, the complexity of full detailed analysis of these tasks, as well as the proper presentation of this part of the study, is beyond the scope of this present report, except to provide an example and to illustrate some overall patterns which were found throughout the data.

Though there was substantial amount of individual variability, and some variability in overall pattern of scores and responses between the different tasks, all five tasks yielded fairly consistent and parallel patterns of response. In general, K and movement scores were low, and form scores were relatively high. There were not a lot of whole responses, but many part whole responses. Neither were there many minor detailed responses, possibly due in part to the context of presentation and the basic lack of privacy that this involved. The frequencies of the scores for all five samples were recorded, and correlation between the tasks is quite high, as shown by the table below:

Table 12-10. Correlations of Scores of Inkblots

Because these correlations conflate the different scoring categories, they mask some important differences of pattern between the five sets of tasks. For example, the Rorschach and Harrower were much stronger than the others in the elicitation of whole responses as well as in the elicitation of high form responses, also in more use of shading in responses, and proportionately greater numbers of human and human object responses. Also, there were no plant responses for the Rorschach and few for the Harrower versions, and fewer shape, shape, abstract and use of space for the latter samples. Correlations were run between the different scoring categories on the five samples to yield fairly strong positive and negative correlations between all but a few of the categories.

To describe these samples in terms of the gross frequency patterns of the scores used, there are high scores in the Animal content in all five samples, and relatively high F scores for form responses, which were good, but not the highest. On the other hand, there were low scores in movement, K, space, and other categories. Because color scores were not a dimension of the black and white tasks, no such scores could be obtained, but the few colors scores that were obtained from the color version of the Harrower indicate fairly low color responses, and mostly pure color or FC type responses.

Frequency patterns for content scores entailed the search for saliency among and between samples, for recurrent content across different pictures. Three of the five samples were compared for content frequencies, #2, the Rorschach and the Harrower 1 because of the high correlations between these different tasks, and also because of important basic differences in response pattern and scores which were noticed between these tasks.

All three samples yielded a great variety of content, content which overlaps as much as 90%.(Rorschach had approximately 224 different items, while task #2 had approximately 350, and the Harrower about 270 items). This gross overlapping of the total ranges of items on the 3 samples thus analyzed should not be overlooked, for it is in this gross sharing of response pattern that a large part of shared cultural patterning might be defined. Though many items occur only once or a few times in each sample, most of the same items with but few exceptions recur repeatedly in more than one sample. This sharing may not be just the consequence of the overlap of the same subjects in all the samples, but also because the same people are responding to different sets of stimuli in similar ways which evoke similar patterns of response across a common range of possibilities.

Though the range of kinds of items was roughly the same between the samples, the distributions of frequencies within this range varied considerably. The variation between the samples is not so much in kind, as in relative frequencies of things elicited. In analyzing the content frequencies of these samples, it is important to take into account the number of different cards in which certain content items occur. If a high frequency score occurs only on one card, this suggests that the image on the card clearly biased the response pattern. On the other hand, lower frequencies of items which occur both across informants and cards are more interesting for their potential. Some low frequency items my recur in a sample as the result of the same one or two people repeating similar responses to different images.

On all three samples, the vast majority of items occurred only once or twice throughout the sample. Of the 350 items of sample #2, approximately 212 items occurred only once or twice (60.5%). (For the Harrower, 210 items of about 270 occurred only once or twice (77.8%), while with the Rorscharch, 150 items of approximately 220 occurred only once or twice (68%). It is of course this great variety of response which renders these techniques valuable in psychoanalysis. Of the remaining items, 33 of 270 of #2 were of frequencies of 3 or 4 (12%). It is impossible to say what level of frequency might be taken as a significant cut off point, but for sample #2 only about 17 items of 270 (6.3%) have a frequency of 5 or greater out of a sample of more than 350 possible responses.

For sample #2, the most frequently occurring item were birds (60 over 9 of 15 cards), followed by eyes (46 over 10 cards), people and legs ( both 33 over 9 cards), butterflies (30 over 6 cards), trees (30 over 6 cards), nonspecific animals (24 over 8 cards), legs (20 over 6 cards). Lower frequency but still salient items are hearts or "love" shapes (18 over 7 cards), bats (16 over 5 cards), dinosaurs (16 over 6 cards), human faces (15 over 8 cards) wolves (14 over 4 cards) and monkeys (14 over 2 cards), the body (14 over 9 cards), bugs or insects (12 over 6 cards), dogs and buttocks or anuses (12 over 5 cards each), and dancing (11 over 2 cards). Other salient items with frequency scores between 5 and 10 are girls, bears, porcupines, kissing, leaves, crocodiles, ladybirds, mouths, human heads, noses, eyebrows, kidneys, bones, crabs, bees, bird beak's, insect feelers, foxes, mice, hands, kangaroos, rats, kites, ducks and turtles.

High frequency items of the Rorschach are people (60 over 9 cards), legs (41 over 6 cards), butterflies (39 over 8 cards), crabs (37 over 2 cards), birds (32 over 5 cards), shoes or boots (30 over 2 cards), eyes (28 over 6 cards ), hands (23 over 4 cards), women (22 over 3 cards), nonspecific animal forms (19 over 10 cards), human heads (18 over 9 cards), faces (17 over five cards), duck heads (17 over 4 cards ), ducks (14 over 3 cards), mouths (13 over 6 cards), bats (13 over 3 cards), hats or caps (12 over four cards) and dogs (15 over three cards). Other items with a frequency score between 5 and 10 were wings, monsters, wolves, eagles, flies, ribbons, men, dancing, feet, dragonflies, cats, caterpillars, hair, tigers, bears, ears, cat's faces, and sea shells.

High frequency scores for the Harrower were butterflies (42 over 5 cards), people (41 over 9 cards, snakes (33 over one card), eyes (23 over 8 cards), bats (23 over 5 cards), birds (20 over 5 cards), fish (17 over 3 cards ), kidneys (14 over 1 card), dancing (13 over 5 cards), men (13 over 4 cards), hats (11 over 4 cards) and women (11 over 1 card). There were overall fewer high frequency scores for the Harrower, though the frequency of total scores was higher than for the Rorschach, in part because of the clearer version of the copies, and in part because subjects tended to give more specific types of responses for items. Other high frequency items with a frequency score between 5 and 10 are bird heads, cat faces, ducks, dogs, dragons, fish, hands, animal faces, rabbit faces, squirrels, tongues, cats, mouths, legs, planes, hearts or "love shapes", wings, cow heads, chickens and penguins.

Correlations between the samples on the basis of five shared most frequent items (butterflies, people, eyes, bats, and birds) shows a high positive correlation (.75) of frequencies between the Harrower and Rorschach, a low positive correlation between the Rorschach and #2 (.175) and a negative correlation between the Harrower and #2 (-.38). Correlations between the item frequencies themselves show the follow pattern of association.

Table 12-12. Correlations between most Frequent Items.

It is difficult to interpret the real meaning of these frequencies and correlations. Sharing at this level of analysis may be in more implicit terms of domains and shared characteristics of location and quality of response pattern to cards. It is presently beyond the scope of this analysis but there are apparent basic cognitive and conceptual domains under which content items can be grouped on the basis of similarity in terms of the response to the same or similar types of figures. For example, bats, birds, butterflies, airplanes and dragon flies and sometimes heart shapes all form a common class of objects. Another such class of objects were dogs, foxes, wolves, cats, lions, and tigers. The categories into which things are implicitly grouped are not as clear cut, and there are many overlaps of shared traits of items. For example, mice and dogs may sometimes occupy different basic domains, but mouse snouts and dog muzzles may form a common membership.

If there is cultural or psychological relativity in response patterns in terms of content, then it seems to be upon a more superficial level of common class membership, for example, adding a tapir to the list of bears, and cows, which might be a common thing for Malaysians to do, but not expected from Americans. Bats and butterflies, though probably also commonly found in American scores, are perhaps more salient in the Malaysian sample if not for the very ubiquity of these creatures in the local context. Another example of this is seeing sea horses, pipe fish, crabs and other sea creatures. Though there is perhaps of modicum of relativity at this level of association, it appears that the larger domains to which these items are grouped may be more robust across cultural or psychological boundaries. The use of content analysis from projective tests in order to elicit information about conceptual domains might be a promising avenue of further research.

At least one author has mentioned the possibility of cultural and even historical variability of "popular" or common responses to the images of the cards. Recording the frequency patterns of such popular responses may have more theoretical import than the establishment of cross-cultural norms for the psycho-diagnostic administration of such techniques. Rather, they may reveal important linkages of culturally shared response patterns at a deeper psychological level of symbolic response than is normally available to other tasks like those previously considered.

Furthermore, the response patterns which they do elicit across a sample of informants may offer revealing information about some aspects of how information and identity are organized and how complex stimuli may be managed on different symbolic levels. In particular they reveal how information is organized in such a way as to be consistent with self and social identities, and how this identity may fluctuate between an idealized and personalized sense of self projected as human and superhuman image, and a more naturalized image of self projected as an animal form. The projection of the self as some half/human and half/animal form--witches, giants, monsters, robots, aliens, dragons, or even as clouds, responses which are typically regarded as "patho-gnomic" in character. The projection of plant forms must be taken as an important variant of the animal and human form which has not be thoroughly analyzed, but bears a parallel resemblance to the use of tree drawings in the HTP drawing tasks. Also the envisioning of "objects" which are either inanimate or abstracted, seems to be another alternate form of such response in which personalized feeling and identification is repressed or for some reason controlled or "neutralized".

What emerges from this patterning is a kind of structured patterning in the way humans may be actually creating and utilizing symbolic content and form to organize information in their environment in meaningful, and predictable ways--the same object may be "visualized" mentally and psychologically in alternate forms--animal, plant, human, object or some inbetween way. The manner in which the stimuli is thus construed and understood will then have consequences in the patterning of response by the individual. It is not difficult to imagine how threatening content that might arouse anxiety on some level will evoke a different pattern of response than the perception of an image which may arouse interest or excitement rather than anxiety.

The process of "zoomorphization" of stimuli, versus the alternative anthropomorphization of the same or comparable stimuli, or possibly the "objectification" or "abstraction" of stimuli, suggests an interesting process of symbolic interpretation of such material that may evoke uncertain or repressed feelings in the subject. Related to this were few action words encountered which created a verb class of items. People dancing were the most frequently encountered actions in all the samples. Thus dancing, drumming, kissing, flying, swimming, running, standing, embracing, holding, touching, shitting, carrying, pointing, shouting, quarreling, fighting, gossiping were the common actions encountered, whether the actors were human or animal. Examples of this include a few sparse statements such as "two lady rat's gossiping", "the hand of Popeye raised to shout", "two animals friendly touch", "two rats holding hands", "animals kissing", "two animals fighting", "two dolphins kissing", "two animals making love", or "crab without legs and with eyes staring at you".

Also relevant in this regard are the "inbetween" type responses which are in some manner half human and half animal. This category includes witches, angels, giants, satan, aliens, teddy bears, dolls, robots, monsters, beasts, dragons, ghosts of "hantus", "fox-men", batman, spiderman, Garfield, superman, and vampires. In these types of responses we can see the use of a preconceived and fairly stereotypical symbolic form being superimposed upon the stimuli, perhaps as a means of mediating and "framing" such possibly conflicting or contradiction causing stimuli in a meaningful way. An example of such a statement to a possibly sexually evocative inkblot is "An evil queen in her glory dressing."

The presentation of the results of these various tasks was not intended for the purpose of ultimately explaining how and why the people of the Jetty are who they are. This kind of analysis was intended as a means of describing on different levels of analysis common patterns of response to different kinds of "symbolic framing" techniques. Theoretical exposition is postponed until the concluding chapter. Ultimately, though, explanation of these phenomena is not well in hand. We do not really have a good idea of how culture is organized in our brains, and how are brains may function to organize our cultural worlds. These tasks are steeped more in questions than in theoretical understanding or explanation.