CHAPTER 3

by Hugh M. Lewis

The elicitation of "Basic Things" from people was achieved by having them mostly rank order pictures of things presented on paper. Colors were also associated with the things by having people choose different colored as they rank ordered the things. Things fell into a few categories--animals, flora and fauna, general shapes and symbols, household things and miscellaneous things. The principle behind these tasks is that cultural orientation, for whatever unknown reasons, will lead us to select and prefer certain kinds of things over others, and this preference pattern will be consistent across groups of people and over time. We cannot guess at the factors which may be behind what would lead us to consistently choose one kind of thing over another. But the pattern of response in such tasks reveal not only such consistency, but also important inventories and dimensions about the cultural experiences of the people doing the rank ordering.

If we lay fifty common objects on a table, all with practically equal chance of being selected over the other, people of different backgrounds should consistently choose some kinds of things over others. This proclivity of our cultural predispositions gives us a heuristic handle for gaining insight into some facets of how culture may be operating in the minds of subjects.

Out of ten objects, the chances that any one object will be chosen, given all things random, is 10%. Out of a hundred objects, it will be one percent. Thus in a discrimination table based on the frequency patterns of a sample of rank orders, the fact that certain objects emerge much more often than can be expected by chance alone, suggests that something other than random order governs the selection of similar objects by a sample of informants. When these frequency patterns recur across several samples, and across several different kinds of tasks, then the evidence becomes especially convincing.

Basic color shapes task entailed the rank ordering from 1 to 10 of a set of 14 basic shapes. The probability that any single object will be chosen at random will be 7.1%. There were two samples for this task (#1, N = 56, #2, N= 63, #3, N = 17.) That means that for the first sample, there were 560 choices spread over a field of 10 x 14 possibilities. The average score for a random response should be 560/140, or 4, and the average number of responses per category of items should be 560/14, or 40. Thus categories in which frequencies were significantly lower, and especially higher, than this, or frequencies of individual rank orders significantly above or below 4, must be regarded as "salient". Of the fourteen shapes, the following were the most frequent in sample #1: the circle (50), the square and hexagon (47 each), the octagon (45), the triangle (44), the horizontal oval (43), parallelogram (42). Of the remaining shapes, the least frequent were the horizontal rectangle (12), the trapezoid (19), the vertical rectangle (30), the rectangle (31), the upside down triangle and the vertical (36 each). Perhaps more revealing, are the saliency patterns of the individual rank order choices, as the distribution of scores among categories for certain items is far above the expected frequency of four. For the first choice is the hexagon (18) and the circle (8) and octagon (7). For the second choice are the circle (10) and the pyramid and octagon (7) each. For the third choice is the square (10) and the pentagon (9). For the fourth choice is the circle and the pyramid (8 each). For the fifth choice is the horizontal (10) and the octagon (9). For the sixth choice is the horizontal (7). For the seventh choice is the parallelogram (10) and the vertical (9). For the eighth choice is the parallelogram (9), for the ninth choice is the trapezoid (8) and for he last choice is the rectangle (8) and the parallelogram (7).

Typical of each significant category is a saliency "peak" which may span the range of 2 or 3 rank orders. These peaks may overlap with one another and some categories may have two peaks. Each category may have a peak, but such peaking is especially salient for those that are more salient overall. It is expected that such peaks should be significantly lower for the last ranks, and for the lower frequency choices.

A theory of basic shapes would suggest that more basic shapes, particularly the square, the circle and the triangle, would be preferred over more irregular shapes such as the trapezoid and the parallelogram. Furthermore, when there is strong saliency of non-basic shapes, then this must be particularly indicative of a culturally salient pattern, and especially when this is strongly a first choice, such as the hexagon.

Sample #2 (N = 63) would expect an average frequency per category of 45 and an average score 4.5 per rank. The most frequent category is again the circle (55) followed by the horizontal (54), the hexagon and the pyramid (53), the square (52), the octagon (50), and the vertical (48). The least frequent shapes of the second sample was the horizontal rectangle (19), the trapezoid (26), the vertical rectangle (37), the parallelogram pentagon and upside down triangle (43) and the finally the rectangle (44). The rank order frequency reveal the hexagon and circle (15 each) for first choice, followed by the square (8). The square was the strongest second choice (12) followed by the hexagon (10) and the octagon (8). The circle was the strongest third choice (13) followed by the triangle (9) and the hexagon (7). The fourth choice was the vertical (9) and the upside down triangle and the hexagon (7 each). Fifth choice included the pentagon and hexagon (8) and the octagon (7).

The sixth choice is the horizontal (11) and the pyramid (10) and the rectangle (7). Seventh choice is the parallelogram (9) and the rectangle (8). The eighth choice is the pyramid (8) and the pentagon and rectangle (7 each). Ninth choice includes the horizontal, vertical and upside down triangle (7 each) and last choice is the parallelogram (13) and the trapezoid (10) and the horizontal rectangle (8). One should expect from this pattern that the basic shapes as strong in the first five, and the irregular shapes more apparent in the last five ranks. One might also expect that the "inbetween" shapes, for instance the rectangle, pentagon, upside-down triangle, and parallelogram, to be the least salient overall and the most spread out between the ranks. The vertical rectangle is the least overall salient shape, with a peak of only 5.

In sample #3 (N =17) one would expect a category frequency of 12 and a rank-item frequency of 1.2. Because this is the smallest sample, one might also expect the greatest departure from the patterns presented above. The most frequent categories are the hexagon and the square (17), followed by the circle and the horizontal (16) and then the pentagon and the pyramid (15), and then the upside down triangle (13). The least frequent categories include the trapezoid (2), the horizontal rectangle (6), the vertical rectangle (10 each), the rectangle (11) and the parallelogram and vertical (12). In terms of rank order, the hexagon is the most frequent first choice (7), followed by the circle (4). For the second choice is the circle (5) and the hexagon (4). Third choice is the square (7). Fourth choice is the pentagon (4). Fifth choice is the octagon and horizontal (4). Sixth choice is the triangle (4). Seventh is the vertical rectangle (5), eight is the rectangle (4). The ninth choice has no salient ranks, except for the horizontal (3). The tenth has the pentagon as the most salient (5). The lowest frequency categories were also the least salient in terms of rank frequencies.

There is a positive correlation of .96 between sample #1 and sample # 2, a positive correlation of .89 between sample #2 and sample # 3, and a positive correlation of .87 between sample #3 and sample #1, showing high consistency in response pattern.

The correlations of gross frequencies between the shape categories for the three samples was also run, revealing an interesting pattern, a .9 + positive correlation for all the different shapes.

It can be expected that a similar pattern holds across the different tasks for basic things. For example in the selection of animals, there were unfortunately four fractured sub-samples (#1, N = 16, #2 N = 23, #3, N = 23, #4, N = 14) due primarily to successive revisions of the original task. Because the sample sizes are smaller, a greater amount of variability of pattern can be expected compared to a larger more robust sample. The similarity between the different versions is close enough, and the overlap between the samples small enough, that they can be readily compared, and even combined without significant conflation of the results. There were seventeen animals which were ranked to 10. In the first sample the expected frequency per category is 9.4 (rank frequency .94). In the second and third samples this expected frequency is 13.5 (rank frequencies 1.35). In the fourth sample it is 8.24 (rank frequency is .824).

For the first sample the most salient categories are shark (15), horse and elephant (13), rooster and fish (12), the tortoise (11) and the cow (10). The least salient categories are the monitor (1), the crab (3), the pig and the dog (6), the snake and the chicak (7) and the crocodile and the mouse (9). The rank order of salience was from first to tenth place: 1. horse and fish (6), 2. fish and rooster (4), 3. rooster (4), 4. the elephant (3), 5. cow and elephant (3), 7. horse (3), 8. the crocodile (4), 9. crocodile (2) and 10. the shark (10).

For the second sample the most salient categories were the horse (22), the tortoise and fish (21), the cow (20) the elephant (18), the rooster (17), the crab (16), and the least salient categories were the monitor (4), the mouse and the crocodile (6), the shark (8), the dog (9) snake (10), bird (12), penguine (13). The rank order in terms of highest frequency was: 1. the horse (15); 2. the fish (10); 3. tortoise and crab (7); 4. fish (5); 5. cow (5); 6. the cow (4); 7. the elephant (4); 8. the rooster and cow (4); 9 the chicak; and 10. the snake (4).

In the third sample most frequent are the cock (24), the horse (21), the cow (20), the fish (18), the elephant (17),the tortoise (16) and the snake (15) and least frequent are the monitor (4), the shark (5), the pig, crab and crocodile (6), the the penguin (9), bird (10), penguin and chicak (11).The rank order is: 1. horse (8) 2. the horse, tortoise and fish (3), 3. the elephant (4), 4. the rooster (4), 5. cow (5), 6. the fish, rooster and bird (3), 7. the snake and rooster (4); 8. the rooster (8), 9. the elephant (2) and 10. the mouse (2). The scores of this sample are a little skewed because a few of the rank orders in it were not complete in the ninth and tenth ranks.

The fourth sample (N =14) the most salient categories were the horse (13), the fish and elephant (12), the penguin (11), the cow (10), the tortoise and crab (9), while the least salient categories were the pig and mouse and dog (4), the rooster, monitor and crocodile (5), the bird (6), the snake (7) and the shark (8). Rank frequencies are: 1. horse (7), 2. penguin (4), 3. none, 4. the cow (4), 5. the elephant (3), 6. shark and mouse (3), 7. none, 8. none. 9. none. 10. none. This last sample suggests that a size of 14 is too small to produce significant patterns, the variability of the sample is too great. On the other hand, the fact that this was a color-ranked task compared to the previous three sub-samples may have something to do with the minor differences in rank frequencies.

The use of color ranking and association was used extensively in the basic items tasks, and magnifies the complexity and scale of the samples by several times. The full analysis of these color patterns was not undertaken. For the preceding sample, there are few clear color saliencies, and this lack of clear association between colors and objects suggests that objects are not associated to colors, but colors are associated with objects. Thus, the selection of the color before the object led to the across the spectrum coloring of different objects. There may be patterns in color combinations with different objects, but this requires more complex analysis. Only several rank item ratings at a score of 3. The horse was associated curiously with both purple (3) and black (3), the crab with red (3), bird with yellow (3), penguin with blue (3). Otherwise most associations were missing.

In terms of rank orders of colors themselves, purple was the most frequently chosen first color (4), pink the second color (3), none for the third rank, purple for the fourth (3), none for the fifth or sixth, blue for the seventh rank, brown for the eighth (5), grey for the ninth (3) and black for the final choice (6). These patterns replicate the frequency patterns of the Luscher color tasks, and can be expected to recur throughout the color-rank basic items tasks. The lack of any clear saliency pattern in the third, fifth and sixth positions may be due either to the small size of the sample, or to the subjects attending to the ranking of objects and colors at the same time, or perhaps also to the inherent lack of association of objects to colors, which made color selection less important than object ranking. There may well be a negative correlation between color rankings and associations and object rankings and associations, such that the saliency of color frequencies will be substantially lower than those for the object rankings.

Another rank order task involved the selection of basic "symbols" (really "signs" such as stars, crosses, etc.) A sample of N = 38 was collected with an average category frequency of 15.2 and an average rank-item frequency of 1.52. The most frequent categories were the five point star (33), the crescent (32), the heart and the mandala (29), the diamond and the eight point star (26), the Greek cross (25), the eight point star (22), the arrow (20), the X-mas tree (18). The least frequent categories were the period sign (0), the sickle (2), the equal sign and the sworl (3), scales (4), the "x" and the cross (6), the flag and the pyramid (7), the fisheye (10), the MAS logo (13) and the swastika (14). Rank order frequencies were: 1. the 5 point star (12) and the heart (6), 2. the mandala (9) and the 6 pt. star (7) and 5 pt. star (5), 3. the five pt. star and the heart (6), 4. the X-mas tree (6) and the mandala and heart (5) 5. the crescent (6), 6. the diamond (7), 7. the 8 pt. star (7) and the crescent, Greek cross and diamond (5), 8. the Greek Cross (7), 9. the flag (10), and 10. the arrow (5).

This sample was corroborated with a smaller overlapping sample (N=17) in which the expected category frequency is 6.8 and the expected rank-item frequency is .68. The most salient items are the 5 pt. star (16) heart and the mandala (15), the six pt. star and the Greek cross (14), the 8 pt. star (13), the crescent and X-mas tree (12), the swastika (10), and the gold and diamond (8). Least frequency categories are the sickle, period, sworl and fisheye (0), the ankh, and scales (1), the equal sign (2),the flag, saw, christian cross, (3) the "x" and the MAS logo (4). Rank order frequencies are: 1., the Mandala (6), 2. 5 pt. star (7), 3. crescent and heart (3), 4. 6 pt. star and heart (4), 5. Greek cross (4), 6. gold sign (6), 7. Greek cross (4), 8. nothing, 9. the heart (3), and 10. the diamond (3).

Another task of basic things involved the 12 order ranking of 37 common household objects. A large sample (N = 88) yields a rank item frequency of 2.38 and a category frequency of 28.5. The most salient categories were the bicycle (70), shoes (65), the television (59), the armchair (56), sunglasses (57), alarm clock (55), the skeleton key (41), the fountain pen (40) open box (34), coffee cup (33) ribbon and ruler (32) and stapler (31). Least salient categories were the scotch tape (1) and the eraser (1), soap (3), the black box and the tube of glue (4) the pushpin and thermometer (7) the photo album (9), the boiling pot with handle (12) the pen knife and the right-angled square and pencil (14) the bird cage (17), the heart and the spanner wrench (20) the postage stamp (23) and calendar book (24). Other items falling in the upper twenties range include the address book, shower head and robot (27). Rank item frequencies were: 1. bicycle (28) and sunglasses (12) and chair (9); 2. bicycle (9), skeleton key and arm chair (6), sunglasses, robot, alarm clock, ribbon (5), 3. the television and sunglasses (8) and the arm chair and calendar (7), 4.the alarm clock (8) and the television, skeleton key and armchair, 5. television (10), shoes (9), bicycle (6) and alarm clock (5), 6. the alarm clock (9), open box (7), sunglasses (6), 7. the arm chair and shoes (6) and the open box (5), 8.fountain pen (7) and the bicycle, pencil and sunglasses (5), 9. the ruler and shoes (6) and the television and the shower head (5), 10. the television (7), sunglasses (6), and the armchair, shoes, postage stamp and coffee cup (5) 11. the television (7) and ruler (5), and 12. shower head and address book (4).

Another task involved the ranking of 12 of 50 of different images of plants and animals or "flora and fauna". Analysis of the sample (N =76, category frequencies equaling 18.24 and rank-item frequencies of 1.52) showed that highest frequencie categories were the crab (52), the snake (42), the dog (40), the crow (39), the second crab (34), the pony and the second sea turtle (33), the starfish and the frog (30), the crocodile and the fish (28), the tyrannosaurus and another frog (27), the squirrel (25), the pine tree (24), the shell (23), the lizard (22) the clam and the snail (21) and the dragonfly (20). It is interesting that only one plant and no insects were represented in the high frequency group, though 21 such forms were presented. Low frequency items were the dinosaur, the beaver cactus (2), the oak leaf, small cactus, ant, and trilobyte (3), seaweed, grasshopper and bee (4), the acorn and fig tree (5), the beetle, flower, coconut palm (6), the spider and pterosaurus (7), the palm tree and the dolphin (9), the second ant (10), the archaeopterix, saguru cactus and a sea turtle (11), the second pine tree, the "magnolia" (12), the bee (14), the quail and the beetle (15), and the mushroom and the tuna (16). Rank order salience include: 1. the crab (14) and tyrannosaurus (9) and second crab (8) and beetle (7), 2. the crab and dog (7), the crow, crab and frog (6); 3. the dog (7), 4. the crow (8) and starfish and the dog (7)m, 5. the crow and the squirrel (6), 6. the dog (7), 8. the pony (6), 9. the seaturtle (7), 10. the starfish and dragon fly (5), 11. none, 12. none.

These different rank order task of basic items were corroborated with a separate and differently constructed rank order task utilizing nine sets of many of the same items but differently arranged on the page and differently configured on a "framed" background and field. Question of the relative bias absolute position of the figures on the page, the relative juxtaposition of the objects in relation to one another, and their relative sizes needed to be dealt with in some manner. The effect of presenting these figures on a more ambiguous background was to render the task more difficult to complete, and to make the problem of rank ordering more prone to error as people would chronically skip over a number or repeat a number twice.

One example of these tasks was of 19 "shapes". The sample (N=21) has a category frequency score of 10 and a rank item score of 1. The most frequent categories were the horizontal oval (21), the circle (20), the hexagon (19), the square, triangle and the octagon (18), the oval (16) the parallelogram (15), the small oval (12), and the pentagon (11) while the least frequent categories were the arrowhead, the chisel, and the right triangle (1), the horizontal rectangle and the diamond (3), the vertical rectangle (4), the trapezoid and the upside down triangle and the rectangle (7). Rank item frequencies were: 1. the circle (12), 2. the horizontal oval (8), 3. the hexagon and circle (4), 4. the small oval and the square (5), 5. the triangle and the octagon (5), 6. the oval (5) and the upside down triangle (4), 7. the parallelogram (4), 8. the triangle (4), 9. the pyramid (4) and 10., the parallelogram (4).

Table 12-10. Correlations between different basic shapes

Correlation between 10 objects from this sample and same objects from the #3 (N =17) and the #1 (N=56) samples presented above shows a positive correlation of .71 between this and #3, .93 between #3 and #1 and .84 between this one and #1, and the following high positive correlations between the different objects as shown on the previous table.