The Tragedy-Art Connection:
Shakespearean Tragedic Language and Artistic Preference
New Critical Light: Final Report
MMAM/ITCHS
Shakespearean Correspondence Assessment
Abstract
In early 2017, the Institute for Travesty, Comedy, and Humor Studies ran an experiment in cooperation with the Minnesota Marine Art Museum. The test’s formal hypothesis was that preference for various Shakespearean rhetorical forms is related to artistic preference among 24 paintings in the museum’s main-gallery collections. This report documents how the test was designed and administered to 42 greeters and security staff at the museum and additionally discusses the 18 high-confidence results which together constitute proof of the test hypothesis: preference for Shakespearean rhetorical forms is related to artistic preference.
Introduction
In February 2017, Robin Jaeckle Grawe, Executive Director of the Institute for Travesty, Comedy, and Humor Studies (ITCHS) met with Nicole Chamberlain-Dupree, Executive Director, and with Heather Casper, Curator of Education, of the Minnesota Marine Art Museum (MMAM) and proposed a joint undertaking with the museum whose formal research purpose would be to determine whether preferences within Shakespearean rhetoric (a totally verbal form of art) had any corresponding relationship to artistic preferences among paintings.
Toward this end, ITCHS volunteered to design an assessment to be taken by volunteer greeters at MMAM. The test instrument would center on a Shakespearean Tragedic Language Assessment (STLA) which ITCHS had previously designed and an artistic assessment whose basic direction would be determined by MMAM staff and then structured for the assessment by ITCHS.
MMAM staff met and arrived at eight painterly categories that were worthy of study given the resources of the museum. For each of these eight painterly categories, the staff nominated a number of paintings that were clearly superlative in this dimension of painterly interest. The quality of the MMAM collection thus becomes centrally important, and ITCHS was convinced that MMAM’s collection of European art with names like Picasso, Monet, Renoir, Signac, Constable, and Dix, its collection of American art with names like Audubon, O’Keefe, Homer, and Knight, and its collection of Hudson River School paintings with names like Cole, Heade, Moran, Bradford, and Quidor was a tremendous resource for testing the comparative appreciation of masterwork paintings.
General Results
The premise of the MMAM/ITCHS study was that there is a relationship between preferences among types of rhetoric employed in Shakespeare’s tragedies and appreciation of paintings in the MMAM collection.
A total of 78 statistical tests were run on the MMAM/ITCHS dataset to test this premise. A total of 18 of these statistical tests resulted in high-confidence (95+% results). A number of results over 99% confident were recorded. The highest confidence was for a test that showed a z-statistic in excess of 5.0. The best z-score table available to ITCHS extends only to a z-score of 4.0, which is 99.994% confident.
While there are obviously 60 tests that did not result in high-confidence results, there is no reason to believe that all of the tests can in theory have high-confidence results for painterly categories’ relationship to Shakespearean rhetoric. That 18 relationships were found is an overwhelming argument in favor of the proposition that artistic preference and Shakespearean rhetorical preference are related.
Painterly Quality Assessment Design
MMAM staff came up with eight painterly qualities. These were condensed by ITCHS staff to six. MMAM staff had suggested superlative individual painting examples for each of their eight categories. ITCHS staff worked almost exclusively with these, to put together six groups of four paintings which respondents would be asked to rank from 1 (most appreciated) to 4 (least appreciated) relative to the others of that group. It was felt that six such groups would challenge respondents but not entirely overtax them in making the fine discriminations involved.
Each of the six preference groups included representatives from four of the six painterly categories. The four choices were limited to a single gallery or from the adjoining Hudson River and American galleries. Care was taken to make paintings as often as possible visible from a single spot within the museum. Thus, for example, the Heade paintings were placed in the same group with de Haas, all of which can be seen from a single position near the doorway between the Hudson River and American galleries. The fourth painting in the group, Arthur Parton’s A Pool in the Adirondacks cannot be seen from the same spot as de Haas, but can be viewed along with the two Heade pieces from many spots.
As the Heade and de Haas paintings also exemplify, care was taken to put paintings of roughly the same size in opposition to each other in the same group. Thus, small masterpieces like Silva’s Sunrise, Bernegat Beach, New Jersey and O’Keefe’s Lake George Autumn were not in direct competition with monumental works like Heade’s Great Florida Sunset and View from the Fern Tree Walk, Jamaica.
European gallery paintings were in direct competition only against other paintings from the European gallery.
The six condensed categories of painterly quality used in the assessment were
Light
Depth
Harmony or Peacefulness
Movement
Chaos or Aggression
Toward Abstraction
Clearly these six are not parallel in grammatical construction, but we represent them so here because this representation reflects the impetus of both staffs in making their choices. Hereafter, we will use a short-hand of parallel construction for naming the six, simplifying Harmony and Peacefulness to Harmony, Chaos or Aggression to Chaos, and Toward Abstraction to Abstraction.
Each of the six painterly elements was represented four times in the six preferential groups of four. Not surprisingly, the six painterly elements were not evenly represented in the three galleries, the European gallery represented in this survey entirely by Depth, Movement, Chaos, and Abstraction.
A full list of paintings used in the six preference groups along with the pre-determined painterly element represented by each is found in Appendix A.
It should be noted that there were many balances within the six painterly categories chosen for this study. Light, Depth, Harmony, and Movement constitute something of a painterly “quadrilateral” as ITCHS has defined “quadrilateral” in numerous book-length studies.
Chaos and Abstraction form their own “wing” of modern, primarily ideological and European approaches to artistic experimentation. Light and Depth are another “wing,” often totally interdependent on one another. Movement and Harmony are both complex qualities dependent on many disparate choices within an artistic work. Harmony and Chaos seem to represent fundamental alternatives. Harmony, Chaos, and Abstraction all seem more “cerebral” compared to Light, Depth, and Movement.
These balances are not necessary to the formal assessment of possible relationship between appreciation of Shakespearean rhetoric and appreciation of artworks. The balances were, however, carefully undertaken and allow the dataset of the formal study to be used for many subsequent studies. Establishing these balances worked to eliminate potential biases that might enter into selection of artistic works. The formal balances are exactly that—balanced—so that, if there is an undetected bias in one category, that bias can be countered by an equally numerous bias in some other category.
Other Artistic Preferences
Three other questions of artistic preference were asked. Each had its own form, chosen to give the museum valuable additional information about its collection.
The most extensive of these questions concerned appreciation of Emmanuel Leutze’s Washington Crossing the Delaware. Respondents were asked what they most appreciated about the painting: its historical values, its provenance, its symbolism, its draw to the museum, and its artistic qualities.
The second question asked respondents to rank their appreciation of the three galleries.
The third question asked respondents to list their three favorite paintings. Not surprisingly, James Hope’s Rainbow Falls, Daniel Knight’s The Brook, and Leutze’s Washington Crossing the Delaware were the most frequent answers.
Side Tests
Had the assessment involved only artistic preferences, ITCHS would still have asked for a series of sociological side tests. The simplest of these are respondent gender identification and age by decade. Additional questions asked for relationship to MMAM (notably greeter, docent, or security staff); possibly being a Friend of Will (a volunteer at the local Great River Shakespeare Festival); possibly having taken a course in art history, art, or Shakespeare; primary academic stance in life; and participation in any of eight named alternative preferences for spending retirement or weekend time.
ITCHS asks for such side tests because in our long involvement with assessment, we often find that such information can make crucial differences for understanding the responses we receive. Happily, a great deal has been clearly established in the present study without resort to differentiation among side-test alternatives. And, therefore, we will not report on results dependent on side tests.
However, just as an example of the power of such tests, it was found that respondents who do gardening as a hobby are different in their artistic appreciation from those who don’t (since an overwhelming majority of the respondents were over 60, this is perhaps better stated as a result about Gardeners over 60 years of age). Specifically, Gardeners were more favorable to Abstract art, controlling for some aspects of Shakespearean rhetoric. Since flower arrangement and garden design both have significant abstract design components, this should hardly be a great surprise, but it is reassuring that the study was powerful enough to show this distinction clearly.
It could also be shown that men and women have contrastive slopes of appreciation between specific Shakespearean preferences and art preferences. Whether one’s academic stance was from the Arts, Humanities, Social Sciences, Medicine, Science or Math, or Business often showed an effect on appreciation.
In short, the dataset allows for exploration of all sorts of additional questions in art appreciation beyond those which have been formally studied in the MMAM/ITCHS initiative.
Personal Codes
At the top of each response form, respondents were asked to insert a Personal Code.
The primary purpose of such a code is to allow all responses to be treated with full anonymity, something ITCHS requires in all its testing programs.
The present final report covers, then, anonymous group results.
However, ITCHS has also prepared individual respondent reports for each respondent including the respondent’s scores for both Shakespearean rhetoric and for painterly qualities. These are being distributed back to respondents in envelopes marked with Personal Codes for respondents to pick up near the volunteer sign-in desk in the museum.
It has been suggested by respondents who have received back their personal scores that a presentation by ITCHS for respondents might be helpful in fully understanding the results. ITCHS agrees with this perception and is willing to make such presentation.
Personal Codes are also important for any on-going assessments at the museum. Such a “second round” is currently in progress. If respondents use the same Personal Code in the second round as in the first, the second dataset can be melded with the first for potentially deeper art appreciation studies. ITCHS personnel have made extensive efforts to help in the remembering of codes. Since most respondents have chosen the last four numbers of their telephone numbers, problems of code memory are considerably simplified.
ITCHS Associates
Well over 6,000 individuals have contributed to various ITCHS assessments in the past. ITCHS has indicated to all of them that they are now considered associates of ITCHS, that ITCHS owes them a debt of gratitude and a certain debt to tell them what the study means for them personally. Each associate is “paying forward,” being educated by things previously discovered through associate response and in turn providing responses that may enlighten new associates in the future.
We at ITCHS take ITCHS associateship very seriously and have repeatedly emphasized and cited respondents’ centrality in numerous published studies, many of them reported internationally. We continue that tradition by recognizing our indebtedness to our volunteer respondents here.
We at ITCHS also consider ourselves allied and associated with organizations that have participated in our studies. We continue that tradition with our thanks to MMAM, to its staff generally, to Heather Casper, Curator of Education, for her coordination of efforts, and particularly to Chris Lennon and John Collison, who as security guards were invaluable in getting volunteer cooperation and in guiding volunteers through the assessment process.
Shakespearean Tragedic Language Assessment Design
For better than a quarter century, ITCHS has been designing academic assessments. The formal title, Institute for Travesty, Comedy, and Humor Studies represents our early efforts and successes in assessment of humor preferences. In the last half decade, however, ITCHS made a major policy decision to begin an investigation of tragedy analogous to its investigations of comedy, which are now represented in many book-length discussions. In that context, ITCHS designed a Shakespearean Tragedic Language Assessment (STLA).
An original pilot of the STLA was so highly successful that ITCHS moved immediately to produce a book of tragedic criticism heavily leaning on the empirical results of the STLA to propose a substantial shift in general tragedic theory since Aristotle and in the appreciation and description of Shakespeare’s tragedic achievement. That academic study, In Search of Shakespearean Tragedy, was published by Lap Lambert Academic Publishing in 2016.
Thus, in proposing an MMAM/ITCHS joint study, ITCHS already possessed the Shakespearean test instrument, an instrument that had already been discussed at length in an academic literary discussion and had been honed and enlarged based on its original success.
For purposes of the MMAM/ITCHS assessment, it is sufficient to understand that the STLA measures four different rhetorical forms for their relative strength to a particular respondent.
The four rhetorical forms assessed in the STLA are:
Aptness
Assessment
Eloquence
Elegance.
In Search of Shakespearean Tragedy thoroughly defines and discusses these four elements in their literary guise. A copy of the STLA instrument used in the MMAM/ITCHS assessment, annotated to indicate which of the four rhetorical forms is at stake in choice of each of the items on the assessment, is provided in Appendix B.
Development of an Assessment Testing Procedure:
Artistic Indices
As already indicated, MMAM staff had indicated the superlative examples within its collection of the six painterly qualities which ITCHS staff molded into testable form. The assessment as formed would have six appreciation-ranking questions, each a set of four paintings. A particular painterly quality would be represented only once in any of these six ranking sets.
The four paintings representing each painterly quality (which were originally ranked against representatives of different painterly qualities in four of the six ranking sets) were added together to form a Painterly Quality Index. For example, four paintings from four of the six sets represented superlative use of Light. Each respondent’s ranking of the four paintings was added together in a personal Light Index Score (which was reported back to respondents individually in their personal reports).
Additionally, ITCHS computed the sum of rank preferences for the entire test group (42 respondents) and then ranked paintings against each other for total preference over the whole group. The two most appreciated paintings by rank from each gallery were then used to create a Favorites Index.
For all six painterly categories, the individual respondent’s raw index score was divided by that respondent’s Favorites Index Score. It is this Adjusted Painterly Index Score (APIS) that contributes to the group results for each of the 78 tests conducted by ITCHS. All 18 high-confidence results are correlations between a Shakespearean rhetorical statistic and an APIS—Adjusted Painterly Index Score—for each respondent in the 42-response sample. Since there are 42 respondents in the study, individual rhetorical/APIS responses can be plotted on a scatter graph. The 42 responses thus form some sort of pattern on the scatter, and this pattern can be mathematically analyzed.
Rhetorical Indices
The STLA is specifically designed to generate four indices representing the respondent’s relative preference for the four forms of rhetoric assessed: Aptness, Assessment, Eloquence, and Elegance. These are the Analytic Indices, and one’s score for each index can be directly translated as a percentage of total awarded appreciation. Thus, if one has scored a 7 for Aptness, meaning that 7 times during the STLA one preferred the Shakespearean quotation scored as Apt over an alternative scored as either Assessing, Eloquent, or Elegant, that score can be divided by 18, the number of preference questions on the test, to give a percentage appreciation score, in this case 38.88%. Since random choices throughout the test would give a 25% appreciation score for each of the four indices (or 4.5 each), 7 is a very elevated score, as 38.88% is elevated over 25%. An Aptness score of 3, conversely would indicate a quite weak appreciation of Aptness, in percentage terms only 16.67% of the 18 responses in the assessment, a score considerably below average (the random expectation).
The main point here is that STLA gives four scores, each of which is of relative preference compared to the other three. Preference is thus key to the assessment in all its mathematics.
Again, there are four Analytic Indices computed for the STLA. But there are six combinations of two indices. If we think about such combinations as “lead” indices, it is possible to describe six synthetic indices of lead pairs.
Additionally, we can subtract one index from another. Again, there are six such combinations, and they all show relative preference of one index over the other.
Our research goal was to test whether artistic preference among painterly qualities is related to Shakespearean rhetorical preferences. At the simple level, we would have succeeded if we proved that artistic preference was related to any one of the four rhetorical Analytic Indices. But we were willing to consider more complex relationships as well, relationship to a “lead pair” of indices (a Synthetic Index) and relationship to a relative preference between two rhetorical preferences.
That means that we were willing to run 16 tests (4 analytic + 6 lead pair + 6 relative preference) for each of six painterly qualities.
However, it turns out that lead pairs, as synthetic combinations, themselves come in pairs, the first having constituents which the second lacks and the second having constituents the first lacks. This reduces the number of tests needed to be run for lead pairs by half. The same, however, is not true for relative preference between two qualities: there are still six combinations of two elements, but there are no opposed pairs.
Thus, subtracting three unnecessary (duplicated in reverse) tests from the lead pairs, we end up with 13 tests for each of 6 painterly qualities or a total of 78 tests to be run and analyzed to test the hypothesis that painterly quality appreciation is related to Shakespearean tragedic language preference. The fact of 78 tests does not suggest that there should be high confidence for any particular proportion of them. The 78 tests merely identify 78 different possible ways that painterly qualities could be related to Shakespearean rhetorical forms. It is theoretically possible that appreciation for painterly qualities would be related to preference for Shakespearean rhetorical form in exactly one and only one way.
Mathematical Relationships
Is it really true that taste is so entirely subjective that there is absolutely nothing that can be demonstrated about how taste is acquired, exercised, or related to the rest of our being?
Statistical mathematics is very powerful to supply objective answers about taste. However, many don’t understand statistics enough to be able to understand what statistics can prove. So please bear with this review of mathematical principles involved.
We will be graphing rhetorical preference on the x-axis. So a low score, like the 3 in Aptness mentioned above, will be toward the left side of the graph. High rhetorical scores like 7 will be to the right of the graph.
The x-axis is the “independent” axis, that is, we are choosing to graph one’s rhetorical preferences as if they dictated how one will like or dislike painterly qualities in art. (Note that this is purely conventional for graphing. In fact, we are making no such assertion that literary taste dictates painterly taste.)
We will be graphing appreciation of painterly qualities on the y-axis. Say a painting quality was very unpopular. It was always ranked fourth in four different ranking sets. That painterly quality’s raw appreciation score would then be 4 x 4 = 16. A highly appreciated painterly quality might be best appreciated in all four sets, and therefore would have a raw appreciation score of 1 x 4 = 4. Unfortunately, that means that a high raw score means low appreciation.
Therefore, we reverse the order of these scores (technically we take the absolute value of 17 minus the raw score), so that the low appreciation score becomes a 1, and the highly appreciated score becomes a 13. This is the first step of creating an APIS or Adjusted Painterly Quality Score. The second step is, as indicated earlier, division by a similarly reversed Favorites Index Score.
A low Adjusted Painterly Appreciation Score is thus graphed low on the y-axis, near the bottom of the graph, where we would expect to find it. A high Adjusted Painterly Appreciation Score is thus graphed high, toward the top of the graph.
We can then graph the rhetorical and painterly appreciation score for each respondent as a point on this graph.
Imagine then a group of points that move in a straight line from the lower left to the upper right. This is said to represent a “positive” correlation: low goes with low, high with high. If we had such a straight line, we would have a perfect match between rhetorical appreciation and painterly appreciation. We say that such a low/low to high/ high relation between x and y variables is a positive correlation.
But then imagine an equally straight line of points, but stretching from top left to right bottom. This would be an equally perfect match (with the same number of data points) but it would prove that liking some rhetorical feature was related to disliking some painterly quality; in other words, we are looking at a negative or “negative-slope” correlation.
If on the other hand, the line were straight across at some y-level, then no relationship would be indicated.
Of course, plots of respondent points are virtually never on a straight line. They are instead above and below any line we can draw. To the extent that the graph looks like a random, more or less circular cloud, statistics will give us what is called a z-score of close to zero, close to no shown relationship, close to the line that just goes straight across.
To the extent that the data points start to resemble a sloped straight line, statistics will give us higher and higher z-scores, the highest z-score being the closest to a sloped straight line. If the straight line thus imitated has a positive slope, the math will show a positive z-score. It will show a negative z-score for a negative slope. But the size of the z-score, negative or positive, will show how good a “fit” we have for some relationship.
That isn’t so hard, is it? Much harder perhaps is a line that is a line but it isn’t a straight line. In particular, what if the line makes a “smiling face” or a “frowning face”?
Let’s start slowly. Say the line is really two lines that meet at a high point in the center of the graph, making a “Christmas tree” or “roof top” (that is, a ^-shape). To the extent that our scatter of respondent points looks like a roof top, for a while, the more one likes a rhetorical feature like Aptness on the x-axis, the more one likes the painterly quality like Light on the y-axis. But somewhere in the middle, a new trend takes over, and from there, the more one likes Apt, the less one likes Light.
Now imagine that the roof top or Christmas tree doesn’t have a pointy top but instead rounds over. Then the interpretation is pretty much the same, except that the new downward trend in the middle doesn’t start abruptly but rather gathers steam over some distance before heading in the opposite direction. In other words, the graph becomes a “frowning face.”
Such graphs are not news in mathematics. The ancient Greeks knew about them. We’ve taken over their word, and we call these configurations “parabolas”.
Some of our 18 high-confidence results are graphed as parabolas. If they are “smiling,” that means that both respondents with very low scores for rhetorical appreciation and people with very high scores for appreciation also have high appreciation for the painterly quality.
In many ways, this paradoxical nature of the rhetoric/art relationship is the great discovery of the MMAM/ITCHS initiative.
Let us go a bit further. We have high z-scores and high confidence scores for 18 measures. That doesn’t at all imply that art appreciation is simply rhetorical appreciation warmed over. But it does say that there is some relationship, some part of art appreciation that depends on something at least similar to rhetorical appreciation. And that, to the best of our knowledge, is a discovery at MMAM using its superb collection that has never been demonstrated previously. Perhaps that relationship is to some third, “middle” factor, that molds some of our appreciation of rhetoric and also some of our appreciation for art. Our graph makes rhetoric the “independent” variable, but it may only be standing proxy on the graph for such a middle, as yet mysterious, factor.
High Confidence and z-Scores.
In the last paragraph, we mentioned high confidence and z-scores in the same breath. Let’s make that clearer also.
The MMAM/ITCHS initiative and all such assessments are actually doing two different statistical tasks using the same data.
The first task is to describe the volunteer respondents as a group. How did our volunteer group, which was composed of 42 individuals, do with respect to Aptness and Light? We produce a graph as described above, and we say, “There, that’s a picture of our group of 42 respondents on the Aptness/Light relationship.”
We are simply describing.
We hope that the museum and the respondents will be interested in that description. As Shakespearean scholars, ITCHS staff are most profoundly interested!
But then, there is a possibility of going on to an entirely separate statistical task.
We may ask ourselves, what is the likelihood not just that our 42 volunteers were like this, but that all groups like this, if they were all tested together, would show this kind of relationship as well. In other words, how high is our confidence that this kind of relationship exists for all such groups as the group we actually sampled?
Saving the mathematical details for a statistics course, the z-score directly indicates the answer. We take the size of the z-score (+2 and -2 are treated exactly alike). We check a z-score table. The table shows for 2 a resultant number of .0228.
What this resultant number means is that, given the data points used, there is a 2.28% chance (.0228) that such a high score actually doesn’t mean anything after all. The result could be just random. By the same token, a random result will seem as high as our result 2.28% of the time.
It would seem that that should mean that we have 100 minus 2.28 or 97.72 percent chance of having discovered the right relationship.
And the seeming is, in fact, the reality, if we have performed a “1-tailed” test. What is that? A 1-tailed test means that before the test we had reason to believe what kind of result we would get. For example, before testing, I’d assume that the men in the group on average would be taller than the women (I’d also assume that the women were in better shape than the men). These are 1-tailed tests. As long as the test result is in the “same direction”—men taller, women in better shape—I am justified in accepting the 1-tailed result at the1-tailed confidence level. But if it turned out that men were shorter than women, my whole way of thinking would have been challenged, and I have no reason to be confident at all.
That then suggests what a “2-tailed” test is about. In a “2-tailed” test, I have no opinion before I begin the study. Does appreciation of Aptness make one appreciate Light as a painterly quality? I have no idea.
So then, when I find a z-score of 2, there’s only a 2.28% chance of getting such a high positive reading, but there is also a 2.28% chance that I could get such a negative reading. So I have 2.28% + 2.28% chances that I could get such a high level of z-score and yet the real truth is that there is no relationship despite what looks so good.
Please note that 2.28 x 2 = 4.56. And 100 - 4.56 is 95.44. So 2-tailed, I still have 95,44% confidence that I have discovered the real relationship for all groups like the one sampled.
Is that good? Well, that too depends.
If I am testing a drug that may have some very bad side effects, I’d like much better proof that it will at least work to cure the disease. For such medical proofs, much better than 99% confidence may be required.
For results in science, like testing mice that have all been raised in the same litter and under identical conditions throughout life, 99% confidence is the gold standard of academe.
But in human-subject research, we can’t get respondents from the same litter, and we can’t even be sure that one did and others didn’t have grapefruit for breakfast. Maybe grapefruit affects taste—for art perhaps or for rhetoric, or even for both in opposite directions.
So in human-subject research, academe accepts 95% confidence as the standard of proof. Our z-score of 2 as a 2-tailed test, then, would be considered proof for the MMAM/ITCHS initiative. (Incidentally, a z-score of 1.96 is the actual cut-off for 95% confidence, 2-tailed.)
Thus, we can consider the MMAM/ITCHS results in two entirely separate ways. 1) They are very descriptive. But, in many cases, 2) they can be seen as proof for all such groups.
Ah, there’s the final catch. What are all groups that are “like” our 42-respondent sample? There can be lots of argument there, and the argument cannot be finally settled by mathematics.
We’d say that all such groups will be predominantly more than 60 years old, will be more heavily female than male, will have some relationship to an art museum and will have had substantial time to be in the presence of great art. As a rough guess, perhaps there are 200,000 people in the United States who could reasonably fulfill these requirements.
We could fight over the 200,000 estimate. Much more important, we should all agree that the group isn’t all Americans or even all American adults. It is probably limited to people in the Midwest, and perhaps to non-metropolitan people.
Despite these potential controversies, it is still true that our result does cover a group of which our respondents are merely a sample. And if that result is high confidence, then there is high confidence for that group that there is a relationship between art and rhetorical appreciation.
So, Jane and Johnny, as you fight with your parents over eating peas or carrots, you will still win the argument with your parents about your idiosyncratic sense of taste, or else they will force you to eat the peas or carrots anyway. But don’t claim to them that there is no way they can demonstrate anything about taste. It’s just your taste they can’t demonstrate.
Finally, the High-confidence results
With this mathematical background, it is possible to present our 18 high-confidence results, representing a variety of suggested relationships between artistic preference and Shakespearean language preference. Those results are summarized with confidence levels in Appendix C. These results prove our hypothesis that preference for Shakespearean rhetorical forms is related to artistic preference.
Moreover, and perhaps as something of a surprise, these 18 high-confidence results show a strong tendency to indicate parabolic relationships between appreciation of rhetorical form and appreciation of painterly quality. As indicated earlier, this additional and unanticipated result is likely to be far more important for the study of artistic appreciation than the comparatively prosaic finding of straight-line (linear) correlations between literary and artistic taste.
All of this deserves further elucidation, but that elucidation is not technically part of reporting the results. Interested readers should continue reading the Afterward for our best current understanding in elucidating these results.
End of Report.
Afterword
Elucidating Mathematical Proofs
Obviously, reference to an appendix of mathematical results should satisfy absolutely no one. Certainly, such a conclusion fails to assess the magnitude of what has been achieved. For readers who would like to understand rather than just know the results, this post-script:
A Straight-Line Regression: Eloquence and Harmony
The artistic quality yielding the greatest number of high-confidence correlations was Harmony. Let us consider first, that there is a relationship between appreciation for Eloquence and an appreciation for the painterly quality which MMAM staff labeled “Harmony” or “Peaceful.” The result had a z-score of 2.05 and a confidence of almost 96% (95.98%).
This is the simplest, straight-line kind of correlation. If a member of the group from which the respondents are a sample is more positively sensitive to Eloquence, that person is likely to be more appreciative of Harmony.
Eloquence is “a beautiful speaking.” In traditional appreciation of fine paintings, those that somehow bespeak Harmony have been routinely highly valued. This is virtually a 1-tailed test.
This test was also simple in the sense that the y-axis had only a single or analytic element, Harmony.
On to Complication: Lead Pair Rhetorical Preferences and Negative Correlation: Aptness + Eloquence and Chaos
Consider then the somewhat more complicated result that Aptness and Eloquence sensitivity combined is straight-line correlated to Chaos, but in this case, the straight line has negative slope. The confidence level is very high. The z-score is 2.645. The confidence level, 2-tailed, is 99.16%.
Aptness “hits the nail on the head.” Eloquence “speaks beautifully.” Put the two of those together. The more members of the group are sensitive to such things (as evidenced by choosing them over alternative rhetorical expressions), the more they don’t appreciate Chaos or Chaotic paintings. Notice that this means Chaotic paintings like those chosen as representative for the present study. Maybe some other types of Chaos exist, and maybe those other types have a different relationship to rhetoric.
But the kind of Chaos we were testing, evidently doesn’t hit the nail on the head or attempt to speak beautifully. Again, this seems almost a 1-tailed test.
Paradoxes and Parabolas: Elegance and Harmony
Linear, very understandable relations between rhetorical preference and appreciation of painterly qualities as demonstrated in selected paintings of MMAM do exist. They are, however, comparatively few.
The comparatively many are complex relationships, parabolic relationships. And again, that may be the great finding from the MMAM/ITCHS initiative. All the parabolas are essentially paradoxical. For the smiling-face parabolas, moving higher in rhetorical appreciation is correlated to moving toward more appreciation of the painterly quality—except that, if you happen to be at the bottom (left side of the graph) of the rhetorical range, you have a very good chance of being quite high in your appreciation of the painterly quality!
Let’s go back to Harmony and to another single-factor rhetorical feature, Elegance: Sensitivity to Elegance is related to appreciation for Harmony. This is a very high-confidence result. The z-score is 3.06. The confidence level is 99.76%.
But the result is not linear. It is instead parabolic with a high point rather middlish on the graph.
Again, what can this mean in English?
Elegance is carefully chosen (and typically expensive, for the carefulness of the choice). Our result indicates that at very low levels of Elegance sensitivity–and at very high levels of Elegance sensitivity—appreciation of Harmony is fairly low. But at middlish levels of Elegance sensitivity, appreciation for Harmony is at its height.
How can this be explained? Our best explanation to date would be that middlish or average levels of Elegance sensitivity is the kind of sensitivity that leads one into the great value of the Harmonious painting and make the Harmonious painting appreciated relative to other great works of art representing other painterly qualities.
Our group of respondents is likely much more sensitive than non-museum-related American adults to painterly values. They’ve been sitting in the presence of those values superbly exemplified. So middlish levels of Elegance sensitivity in context is no trivial matter. But it isn’t an excessive, obsessive, or highly concentrated matter either. Don’t be particularly sensitive, just MMAM-volunteer normal sensitive to Elegance, and you are very likely to like Harmony quite a bit (if you are a member of the group sampled).
Again some other group, say a group of fine art investors, might not exhibit this relationship. Our result is only itself and, with the stated level of confidence, a result about the larger group of which are respondents are a sample.
Complex Parabolic Results: Aptness + Assessment and Harmony
At even higher levels of complexity, most of us should have been glad to review the math basics involved, because we now have to deal with parabolic results but also results that are complex, that involve more than one analytic rhetorical quality.
Let us stick with Harmony examples. There is a very-high-confidence result that combined (Synthetic) Apt and Assessment sensitivity is related to Harmony. The z-score is 2.53. The confidence level is 98.36%
The relationship is parabolic with high appreciation of Harmony toward the middle of the Synthetic rhetorical range.
Put in words, extremely low sensitivity to the combination of Apt and Assessment and extremely high sensitivity to the same for the sampled group correlates with low appreciation of Harmony. In other words, we are confronting a frowning-face parabola. Again, a middlish sensitivity to Apt plus Assessment correlates to high appreciation of Harmony.
Our best interpretation of this is congruent to the one we just finished. Middlish levels of such sensitivity, which for the group are probably substantially higher levels than middlish for most Americans, are the levels of sensitivity which guide one through to the values of the representative Harmonious paintings (if one is a member of the sampled group).
A Surprising Tautology
The result we have just discussed is about the Apt/Assessment combination.
It was mentioned much earlier that there are six such combinations, but we ran only three tests because each combination is one of a pair of mathematical opposites. That now becomes important. The result we cited with 2.53 z-score and confidence of 98.36% confidence is also the result of a second and opposite test about the Eloquence/Elegance combination.
A middlish Apt/Association score implies a middlish (though mirror opposite away from the center of middlishness) Eloquence/Elegance score. A very high Apt/Association score is the same as a very low Eloquence/Elegance score. And a very low Apt/Association score is the same as a very high Eloquence/Elegance score. So the curves of the two tests look the same (but mirror-reversed). Middlish scores are the high scores in both cases.
We conclude then that Harmonious paintings ask for a balanced sensitivity to all four of the rhetorical values tested by the STLA, at least balanced within pairs, one pair being characterized by sensitivity to thought (Aptness and Assessment) and one pair characterized by sensitivity to beauty (Eloquence and Elegance).
To the Same Conclusion: Contrasting Artistic Sensitivities: Assessment - Elegance
And finally—yes finally—let’s consider that we have a parabolic result that Assessment minus Elegance sensitivity is related to Harmony. Now there is a mind-bender!
It is also very-high-confidence, with z-score of 2.9 and confidence level of 99.62%.
People of the group sampled, with very low or very high Assessment-minus-Elegance sensitivity are likely not to find their way into the values of Harmony as opposed to the other painterly qualities against which Harmony was tested. On the other hand, middlish Assessment-minus-Elegance is associated with appreciation for Harmony.
So what is Assessment minus Elegance? It is a measure of sensitivity for Assessment above or below one’s sensitivity to Elegance. It is sensitivity to Assessment relative to sensitivity to Elegance.
People of the group sampled who are very low in Assessment compared to their levels of Elegance sensitivity don’t typically find their way into the value of Harmonious paintings compared to how well they find other painterly qualities. Middlish strength of Assessment sensibility relative to strength of Elegance sensibility for this group is a way to appreciation of Harmony.
Assessment is deep-level thought coming to a summational conclusion. Elegance is a deep-level perception of beauty as representing careful choice. Again, the result suggests a balancing as important to finding the value of Harmony, and now, this balancing is a balance between deep thought and deep appreciation of beauty.
So there it is, as exemplified and explained in the Afterword. There it is in Appendix C as first-in-history demonstrations respecting taste.
We at ITCHS are grateful to MMAM for allowing us to be part of this experiment, simply for the results that affect our understanding as literary scholars. Thanks again to MMAM and to all our wonderful associates who made these discoveries for us, for MMAM, and quite possibly for the history of art criticism.
Appendix A
Paintings Assessed by Gallery
with Relative Appreciation Rank and Notable Painterly Quality
Gallery
Rank
Painting
Quality
Hudson River
1
Martin Heade, Great Florida Sunset
Light
Hudson River
21
Martin Heade, View from the Fern Tree Walk, Jamaica
Depth
Hudson River
14
Arthur Parton, A Pool in the Adirondacks
Harmony
American
12
William de Haas, White Island Light – Isle of Shoals
Movement
Hudson River
9
Hermann Herzog, Alpine Lake Scene
Depth
Hudson River
11
Robert Duncanson, Mount Orford
Harmony
Hudson River
18
William Bradford, Arctic Sunset with Rainbow
Light
American
16
William Bradford, Ships off a Northern Headland
Movement
American
8
Winslow Homer, Winding Line
Harmony
American
17
Jamie Wyeth, The Warning
Chaos
American
20
Georgia O’Keefe, Lake George Autumn
Abstraction
American
7
Edward Moran, Sandy Hook Lighthouse
Light
American
10
John Audubon, White-Headed Eagle
Chaos
American
1
Francis Silva, Sunrise, Barnegat Beach, New Jersey
Light
American
23
Stuart Davis, Anchors
Abstraction
American
13
George Bellows, Paradise Point
Harmony
European
22
Georges Seurat, Seated Man
Abstraction
European
19
Pablo Picasso, The Arena before the Rhone
Chaos
European
5
Edgar Degas, The Trainers
Movement
European
6
Claude Monet, Near Fécamp
Depth
European
4
Alfred Sisley, A Gust of Wind
Movement
European
24
André Derain, The Repose
Abstraction
European
3
Otto Dix, Spring on the Hori overlooking Steckborn
Depth
European
14
Henri Matisse, Port of Marseille
Chaos
Appendix B
Shakespearean Tragedic Language Assessment
With Noted Tragedic Language Form
©Institute for Travesty, Comedy, and Humor Studies 2015
For each of the following numbered questions, decide which of the pair, A or B, you think more grabs you as powerful, and ON THE ANSWER SHEET circle A or B to indicate that preference.
1 A. On mixed emotional response: Assessment
“It seem’d she was a queen
Over her passions, who, most rebel-like,
Sought to be king o’er her.”
B. Amidst the storm: Aptness
“things that love night
Love not such nights as these;”
2 A. Funeral elegy: Eloquence
“I come to bury Caesar, not to praise him.
The evil that men do lives on after them;
The good is oft interred with their bones. . . .”
B. Arguing for prompt action: Elegance
“There is a tide in the affairs of men,
Which taken at the flood leads on to fortune;
Omitted, all the voyage of their life
Is bound in shallows and miseries.”
3 A. Refusing to be provoked into combat: Aptness
“Were it my cue to fight, I should have known it
Without a prompter.”
B. Intervening in a brawl: Eloquence
“Keep up your bright swords, for the dew will rust them.”
4 A. On constancy: Elegance
“The skies are painted with unnumber’d sparks,
They are all fire and every one doth shine,
But there’s but one in all doth hold his place.”
B. Men sometimes are masters of their fates: Assessment
“The fault . . . is not in our stars
But in ourselves, that we are underlings.”
5 A. Remorselessness: Aptness
“things without all remedy
Should be without regard. What’s done is done.”
B. A tragic future: Eloquence
“Tomorrow, and tomorrow, and tomorrow
Creeps in this petty pace from day to day
To the last syllable of recorded time.”
6 A. A prince about despicable work: Assessment
“Tis now the very witching hour of night.
When church yards yawn and hell itself breathes out
Contagion to this world . . .”
B. Princely farewell: Elegance
“Now cracks a noble heart. Good night, sweet prince,
And flights of angels sing thee to thy rest!”
7 A. Tedium: Elegance
“So tedious is this day
As is the night before some festival
To an impatient child that hath new robes
And may not wear them.”
B. Aged sleeplessness, youthful sleep: Assessment
“Care keeps his watch in every old man’s eyes,
And where care lodges sleep will never lie;
But where unbruised youth with unstuffed brain
Doth couch his limbs, there golden sleep doth reign.”
8 A. Benediction: Eloquence
“Good night, good night! Parting is such sweet sorrow
That I shall say good night till it be morrow.”
B. Nominalism: Aptness
“What’s in a name? That which we call a rose
By any other word would smell as sweet . . . .”
9 A. Contemplating murder at night: Aptness
“Put out the light, and then put out the light.”
B. A public man made ridiculous: Elegance
“But alas, to make me
A fixed figure for the time of scorn
To point his slow, unmoving finger at!”
10 A. Witches’ incantation: Eloquence
“Double, double, toil and trouble;
Fire burn, and caldron bubble.”
B. Life: Elegance
“It is a tale
Told by an idiot, full of sound and fury,
Signifying nothing.”
11 A. The perception of wisdom: Assessment
“Wisdom and goodness to the vile seem vile;
Filths savor but themselves.”
B. The indifference of pagan gods: Eloquence
“As flies to wanton boys are we to the gods;
They kill us for their sport.”
12 A. An army captain speaking of his mission’s objective: Assessment
“Truly to speak and with no addition,
We go to gain a little patch of ground
That has in it no profit but the name.
To pay five ducats, five, I would not farm it.”
B. On false friends: Aptness
“Whom I will trust as adders fanged.”
13 A. A final tribute: Elegance
“His life was gentle, and the elements
So mix’d in him that Nature might stand up
And say to all the world, ‘this was a man.’”
B. Ironic comment: Aptness
“Ambition should be made of sterner stuff.”
14 A. On rash lovers’ vows: Assessment
“O swear not by the moon, th’ inconstant moon,
That monthly changes in her circled orb,
Lest that thy love prove likewise variable.”
B. A lover’s conceit: Eloquence
“Two of the fairest stars in all of the heaven,
Having some business, do entreat her eyes
To twinkle in their spheres till they return.”
15 A. On the human condition: Assessment
“Humanity must perforce prey on itself,
Like monsters of the deep.”
B. Resolved to accept suffering: Elegance
“henceforth, I’ll bear
Affliction till it do cry out itself ‘Enough, enough,” and die.”
16 A. Farewell bouquet to a star-crossed lover: Aptness
“Sweets to the sweet: farewell.”
B. A prince’s final request to his friend: Eloquence
“Absent thee from felicity a while,
And in this harsh world draw thy breath in pain,
To tell my story.”
17 A. If doing is all there is to it: Aptness
“If it were done when ‘tis done, then ‘twere well
It were done quickly.”
B. Appropriate courage: Assessment
“I dare do all that may become a man;
Who dares do more is none.”
18 A. A soldier’s farewell to arms: Eloquence
“O now, for ever
Farewell the tranquil mind, farewell content!
` Farewell the plumed troop, and the big wars,
That make ambition virtue!”
B. Kissing his murdered victim: Elegance
“no way but this;
Killing myself to die upon a kiss.”
Appendix C:
High-Confidence Correlations between Shakespearean Rhetorical Preference
and Adjusted Painterly Index Appreciation
Rhetorical Type *
Painterly Quality (PQ)
Curve Shape
Rhetorical Range for High PQ Appreciation
z-score **
Confi-dence **
S-Q
Light
Parabolic
v. low & v. high
2.40
.9836
S
Depth
Parabolic
v. low & v. high
2.29
.9780
A+G
Depth
Parabolic
v. low & v. high
2.73
.9986
A-S
Depth
Parabolic
v. low & v. high
2.67
.9924
Q
Harmony
Straight
high
2.05
.9598
G
Harmony
Parabolic
middlish
3.06
.9976
A+S
Harmony
Parabolic
middlish
2.53
.9886
S-G
Harmony
Parabolic
middlish
2.90
.9962
S-Q
Harmony
Straight
high
2.48
.9868
A-S
Movement
Parabolic
v. low & v. high
2.20
.9720
S-Q
Chaos
Parabolic
v. low & v. high
3.08
.9980
S+G
Chaos
Straight
high
2.64
.9920
A-S
Chaos
Parabolic
v. low & v. high
2.48
.9868
A-Q
Chaos
Straight
high
2.07
.9610
S
Abstraction
Parabolic
v. low & v. high
5.15
>.9999
A+S
Abstraction
Parabolic
v. low & v. high
2.42
.9843
A-S
Abstraction
Parabolic
v. low & v. high
2.73
.9936
S-Q
Abstraction
Parabolic
v. low & v. high
3.07
.9978
* * *
A follow-up experiment in humor and art divided paintings by social aspects in the subject treated: Nature, Ships, Boats, Individuals or duos, Groups of People, and Civilization. It then looked for correlations in preference for four Vitalist humor forms: Potential, Tenacity, Performance, and Creativity. This investigation of artistic social preferences and Vitalist Humor preferences found high-confidence results showing correlations between artistic preference and Vitalist humor preference. See
Final Report of MMAM-2/ITCHS Vitalist Humor Assessment