What do you think?
Rate this book
Clockwork Muse
The Clockwork The Predictability of Artistic Colin The Clockwork The Predictability of Artistic Basic FIRST First Edition, First Printing. Not price-clipped. Published by Basic Books, 1990. Octavo. Hardcover. Book is like new. Dust jacket is like new. 100% positive feedback. 30 day money back guarantee. NEXT DAY SHIPPING! Excellent customer service. Please email with any questions. All books packed carefully and ship with free delivery confirmation/tracking. All books come with free bookmarks. Ships from Sag Harbor, New York. Seller 304717 Art We Buy Books! Collections - Libraries - Estates - Individual Titles. Message us if you have books to sell!
432 pages, Hardcover
First published January 1, 1990
3 people are currently reading
51 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 of 1 review
March 28, 2025
I Remember discussing the principles outlined here with some friends in Paris, about 1979. They were very dismissive about the idea that one could come up with an equation for a painting. My recollection was that I had read about the sort of work that Colin Martindale outlines in this book. Maybe, it was in Scientific American or something similar. But I was rather surprised when I have actually come to review this book to realise that it was only published in 1990. I had actually tried something like this sort of analysis with fairy stories many years before. It seemed that there was a lot of formulaic themes in Fairy stories: things came in threes, stepmothers were inevitably bad news, princesses were beautiful and princes handsome etc etc. But my symbolic logic for analysing stories on a random basis became just too complex. So I was delighted to find that Martindale had actually done the equivalent using computer analysis of words and other statistical techniques......and had actually come up with a theory. Basically, it is an evolutionary theory....that arts forms get more complex with time as we constantly seek novelty. However, it is not quite as straightforward as that. He says:
“If the evolutionary theory is valid, several general predictions may be made about any series of artistic products produced within a given tradition:......
• Measures of arousal potential such as novelty, complexity, and variability should increase monotonically over time.
• Measures of primordial content should exhibit cycles of increasing and decreasing density of words or images or sounds indicative of primordial thought. Periods when primordial content decreases should coincide with periods of stylistic change.
• Across very long periods of time, primordial content should increase. (I diagram these predictions in figure 2.3.)
These predictions hold only if the autonomy of the artistic subculture and the chronic arousal level of a society have remained relatively constant..... The evolutionary theory can be construed in two ways. The weak version is that the theory explains a bit, but perhaps not much, about art history. The strong version is that the theory explains the main trends in art history.
Figure 23 is a graphic summary of theoretical predictions: in cases to which the theory applies, arousal potential should increase across time; primordial content and stylistic change should show long-term increases with superimposed oscillations. Note that primordial content and stylistic change should be out of phase: that is, the former should increase when the latter decreases, and vice versa”.
It would not be in the Arts sphere if these sort of proposals were not met with vigorous rejection and counter arguments. For example:.....A humanistic complaint about evolutionary theory is that it reduces art history to a meaningless quest for novelty. This is rather like rejecting Darwinian theory because it reduces human history to a meaningless struggle for survival...... The theory of aesthetic evolution involves no assertion that artists are motivated solely by a quest for novelty. Artists are interested in accomplishing many other things besides making their works novel.
What these other things are varies unsystematically, whereas the pressure for novelty is constant and consistent. Thus, only it can produce systematic trends in artistic form and content.
In the rest of this book, Martindale presents quantitative evidence supportive of the theory outlined in this chapter. In chapters 3 to 5, I deal with French, British, and American literature. The evidence presented in these chapters was derived mainly from computerized content analyses, which involves using computer programs to count words of different sorts.
In chapter 6, he describes data concerning the history of music and the visual arts. Most of the data in this chapter are derived from ratings by naive observers in order to ensure objectivity. Since he “rather likes his own theory”, it would not do to simply to give his own impressions of what happened in the history of art. He would have an unconscious tendency to twist the data to fit the theory. Unfortunately, most art historians, though well aware that this sort of distortion is a danger, do nothing to avoid it. In chapter 7, he deals, again in a quantitative manner, with the question of whether the arts evolve in synchrony. The evolutionary theory involves the assertion that the main forces of aesthetic evolution are intra-artistic, that art does not reflect society to any appreciable extent.
In chapter 8, he examines the evidence for this assertion. And ascribe a good bit of art history to the direct and indirect effects of a continual pressure for novelty— a pressure that operates on individual artists as well as across great stretches of time. In chapter 9, he examines how pressure for novelty shapes individual artistic careers and even trends within the course of individual works.
Martindale says the term “explanation” means something different. Science searches for abstract regularities or laws. Once found, these regularities are expressed in mathematical equations in order to describe the regularity exactly. ....... It bothers some people no end when I take an exactly analogous approach to poetry. For example, I argue that the impact value or arousal potential of poetry increases across time because of need for novelty. To find out, I construct measures of arousal potential and apply them to series of poetic texts. Then I want to know the exact rate of increase. If I want to know it at all precisely, I need to fit an equation to the data points.
To the extent that the equation fits the data, I say that I have explained such-and-such percentage of overall variability in arousal potential....... seriously, to complain that the equation does not help us understand the meaning of such and such a poem, or that it does not deepen our understanding of the poet's "fundamental mission" or some such thing, is true but silly. [I must say that I agree with this]. Since I haven't undertaken to explain such things in the first place.
the goal of a scientific approach to literary history is ultimately both a set of equations that describe the main trends and an explanation of why these equations rather than others were found. Unless we gather quantitative data and then reduce these to specific equations, what we can say will be so vague as to be not worth saying. The point of statistics and equations is to simplify things, not to complicate them—a result possible, though, only if the reader understands the statistics.
Early nineteenth-century French poetry is romantic. W. T. Jones (1961) enumerates six "axes of bias" that differentiate the eighteenth-century enlightenment syndrome from what he calls the romantic syndrome (the romantic biases are italicized):
1. Order versus Disorder: System, clarity, structure, analysis versus com-plexity, fluidity, disorder, novelty, chance, and indeterminacy.
2. Static versus Dynamic: Calm versus change and frenzy.
3. Discreteness versus Continuity: Either-or division and pluralism versus unity, synthesis, inclusion, pantheism, and monism.
4. Outer versus Inner: Objectivity versus experience.
5. Sharp-Focus versus Soft-Focus: Clarity and distinctness versus threshold phenomena and the ineffable.
6. This World versus Other-World: The here-and-now versus spatial, temporal, or fantasy escape.
This list makes it clear that the enlightenment syndrome corresponds to conceptual thought, and the romantic syndrome to primordial thought.
The data support the implication that poets [British 1550-1949] increase arousal potential (AP) by a combination of primordial cognition (PC) and stylistic change
(SC). The equation AP = 0.58PC + 0.25 SC explains 88 percent of the variation in arousal potential (as measured by the Composite Variability Index) for the periods in question..... This is the same percentage of variation explained by an equation relating arousal potential to the mere passage of time. In fact, we can explain 95 percent of the variation in arousal potential if we combine primordial content and the proportion of types added in a period (A %)
He tested the evolutionary theory's prediction that the amount of primordial content in a given period is related to that in prior periods [for Italian painters 1330-1729]. The autoregressive analysis indicated that the best prediction equation for primordial content in period t (PC1) was PCt = 0.40 PCt-1 – 0.41 PCt-3
This significant fit accounts for 54 percent of the variation, or a little over half of the observed interperiod variation in primordial content. The cycles coincide with generally recognized styles: late gothic (periods 1-5), renaissance-mannerist (periods 6-10), baroque (periods 11-17), and rococo (periods 18-20).
From the late seventeenth century until the middle of the nineteenth century, New England gravestones were carved on flat slabs of native stone (usually slate, but sometimes sandstone, schist, and granite...... Within each general style, there is a good deal of variation in the manner in which the subject matter is executed...... Though I have done no quantitative work on the topic, I have wandered through virtually every old graveyard I have found close to hand.....There is clearly a linear to painterly progression within both the skull-and-wings and the tree-and-urn styles.......... The movement from linear to painterly is a retrogression from conceptual to primordial. Furthermore, within both the linear and the painterly styles, there is an analogous conceptual to primordial movement. It would be wrong to ascribe this stylistic progression to gradual development of skill in stone carving. The late winged-head relief style would seem to require more rather than less skill than the early linear neoclassic style....... Thus, it is gratifying that the evolutionary theory seems to work so well in this context. As I have remarked, a selection pressure has to be long lasting but not necessarily intense in order to work its ways.
[In order to examine whether art styles cross media-form boundaries we designed an] experiment that dealt with examples of baroque, neoclassic, and romantic styles, the three most commonly used cross-media stylistic terms..... Subjects could not correctly identify the artists or the titles of any of the paintings, poetry, or music............ Statistical analysis showed that subjects were grouping together instances of the same style from different media with accuracy far beyond chance. Fifty-five percent of baroque art works, 60 percent of neoclassical art works, and 65 percent of romantic art works were correctly classified. We would expect only 33 percent in each case....... The results of the experiment showed that artistically naïve adults are sensitive to cross-media styles. Presumably, the basis for this sensitivity is an untrained "direct" perception of cross-media similarities in the works of art.
To investigate the ability of children to discriminate cross-media artistic styles, we undertook a partial replication of our experiment with nursery school children having an average age of four years...... They correctly classified 8o percent of baroque, 40 percent of neoclassic, and 60 percent of romantic artworks. The children's sorting was more influenced by medium than was the adults', and the children used one rather than two dimensions to discriminate the styles.
I recall visiting an exhibition on Julio Gonzales in Madrid, about 1978 and it said something about his influence on Picasso and vice versa...in the 1920’s and 1930’s. And as I’ve recently visited the Picasso museum in Barcelona and seen some of Picasso’s very early (realist) paintings, I was struck by the apparent evolution of Picasso’s work. And, sure enough Martindale finds the same thing in his equations] ......An analysis of variance of scores grouped into five-year periods showed significant inter-period differences in both arousal potential and primordial content. The trend in arousal potential is best accounted for by a simple linear increase-in conformity with theoretical expectations. On the other hand, variations in primordial content were not due simply to a monotonic increase, though this increase accounts for 65 percent of the variation in the means shown in figure 9.9. When this trend is removed, a significant autocorrelation at a lag of two periods accounts for 25 percent of the remaining variance. [My impression is that he is pushing his statistics a bit far here and really doesn’t have enough data points with Picasso to make these claims]. Thus, we again find microstylistic oscillations. ........The microstylistic oscillations correspond to fairly clear stylistic changes. The increase in primordial content culminating in period 4 (1910-14) corresponds to Picasso's initial cubist style. The subsequent decline corresponds to his movement toward a more realistic and modelled as opposed to fat and fragmented.
My overall take on the book. I am quite fascinated by it. I really think he has identified something important but I’m not sure that the diversity in the arts is as easily corralled as Martindale would like to claim. In general, I think he has done a pretty reasonable job in his selection and randomisation of his data but, as in the case of Picasso, I think he is often pushing the explanations very hard to fit his overall theory and the data he has accumulated.
Still quite original and really interesting. Five stars from me.
“If the evolutionary theory is valid, several general predictions may be made about any series of artistic products produced within a given tradition:......
• Measures of arousal potential such as novelty, complexity, and variability should increase monotonically over time.
• Measures of primordial content should exhibit cycles of increasing and decreasing density of words or images or sounds indicative of primordial thought. Periods when primordial content decreases should coincide with periods of stylistic change.
• Across very long periods of time, primordial content should increase. (I diagram these predictions in figure 2.3.)
These predictions hold only if the autonomy of the artistic subculture and the chronic arousal level of a society have remained relatively constant..... The evolutionary theory can be construed in two ways. The weak version is that the theory explains a bit, but perhaps not much, about art history. The strong version is that the theory explains the main trends in art history.
Figure 23 is a graphic summary of theoretical predictions: in cases to which the theory applies, arousal potential should increase across time; primordial content and stylistic change should show long-term increases with superimposed oscillations. Note that primordial content and stylistic change should be out of phase: that is, the former should increase when the latter decreases, and vice versa”.
It would not be in the Arts sphere if these sort of proposals were not met with vigorous rejection and counter arguments. For example:.....A humanistic complaint about evolutionary theory is that it reduces art history to a meaningless quest for novelty. This is rather like rejecting Darwinian theory because it reduces human history to a meaningless struggle for survival...... The theory of aesthetic evolution involves no assertion that artists are motivated solely by a quest for novelty. Artists are interested in accomplishing many other things besides making their works novel.
What these other things are varies unsystematically, whereas the pressure for novelty is constant and consistent. Thus, only it can produce systematic trends in artistic form and content.
In the rest of this book, Martindale presents quantitative evidence supportive of the theory outlined in this chapter. In chapters 3 to 5, I deal with French, British, and American literature. The evidence presented in these chapters was derived mainly from computerized content analyses, which involves using computer programs to count words of different sorts.
In chapter 6, he describes data concerning the history of music and the visual arts. Most of the data in this chapter are derived from ratings by naive observers in order to ensure objectivity. Since he “rather likes his own theory”, it would not do to simply to give his own impressions of what happened in the history of art. He would have an unconscious tendency to twist the data to fit the theory. Unfortunately, most art historians, though well aware that this sort of distortion is a danger, do nothing to avoid it. In chapter 7, he deals, again in a quantitative manner, with the question of whether the arts evolve in synchrony. The evolutionary theory involves the assertion that the main forces of aesthetic evolution are intra-artistic, that art does not reflect society to any appreciable extent.
In chapter 8, he examines the evidence for this assertion. And ascribe a good bit of art history to the direct and indirect effects of a continual pressure for novelty— a pressure that operates on individual artists as well as across great stretches of time. In chapter 9, he examines how pressure for novelty shapes individual artistic careers and even trends within the course of individual works.
Martindale says the term “explanation” means something different. Science searches for abstract regularities or laws. Once found, these regularities are expressed in mathematical equations in order to describe the regularity exactly. ....... It bothers some people no end when I take an exactly analogous approach to poetry. For example, I argue that the impact value or arousal potential of poetry increases across time because of need for novelty. To find out, I construct measures of arousal potential and apply them to series of poetic texts. Then I want to know the exact rate of increase. If I want to know it at all precisely, I need to fit an equation to the data points.
To the extent that the equation fits the data, I say that I have explained such-and-such percentage of overall variability in arousal potential....... seriously, to complain that the equation does not help us understand the meaning of such and such a poem, or that it does not deepen our understanding of the poet's "fundamental mission" or some such thing, is true but silly. [I must say that I agree with this]. Since I haven't undertaken to explain such things in the first place.
the goal of a scientific approach to literary history is ultimately both a set of equations that describe the main trends and an explanation of why these equations rather than others were found. Unless we gather quantitative data and then reduce these to specific equations, what we can say will be so vague as to be not worth saying. The point of statistics and equations is to simplify things, not to complicate them—a result possible, though, only if the reader understands the statistics.
Early nineteenth-century French poetry is romantic. W. T. Jones (1961) enumerates six "axes of bias" that differentiate the eighteenth-century enlightenment syndrome from what he calls the romantic syndrome (the romantic biases are italicized):
1. Order versus Disorder: System, clarity, structure, analysis versus com-plexity, fluidity, disorder, novelty, chance, and indeterminacy.
2. Static versus Dynamic: Calm versus change and frenzy.
3. Discreteness versus Continuity: Either-or division and pluralism versus unity, synthesis, inclusion, pantheism, and monism.
4. Outer versus Inner: Objectivity versus experience.
5. Sharp-Focus versus Soft-Focus: Clarity and distinctness versus threshold phenomena and the ineffable.
6. This World versus Other-World: The here-and-now versus spatial, temporal, or fantasy escape.
This list makes it clear that the enlightenment syndrome corresponds to conceptual thought, and the romantic syndrome to primordial thought.
The data support the implication that poets [British 1550-1949] increase arousal potential (AP) by a combination of primordial cognition (PC) and stylistic change
(SC). The equation AP = 0.58PC + 0.25 SC explains 88 percent of the variation in arousal potential (as measured by the Composite Variability Index) for the periods in question..... This is the same percentage of variation explained by an equation relating arousal potential to the mere passage of time. In fact, we can explain 95 percent of the variation in arousal potential if we combine primordial content and the proportion of types added in a period (A %)
He tested the evolutionary theory's prediction that the amount of primordial content in a given period is related to that in prior periods [for Italian painters 1330-1729]. The autoregressive analysis indicated that the best prediction equation for primordial content in period t (PC1) was PCt = 0.40 PCt-1 – 0.41 PCt-3
This significant fit accounts for 54 percent of the variation, or a little over half of the observed interperiod variation in primordial content. The cycles coincide with generally recognized styles: late gothic (periods 1-5), renaissance-mannerist (periods 6-10), baroque (periods 11-17), and rococo (periods 18-20).
From the late seventeenth century until the middle of the nineteenth century, New England gravestones were carved on flat slabs of native stone (usually slate, but sometimes sandstone, schist, and granite...... Within each general style, there is a good deal of variation in the manner in which the subject matter is executed...... Though I have done no quantitative work on the topic, I have wandered through virtually every old graveyard I have found close to hand.....There is clearly a linear to painterly progression within both the skull-and-wings and the tree-and-urn styles.......... The movement from linear to painterly is a retrogression from conceptual to primordial. Furthermore, within both the linear and the painterly styles, there is an analogous conceptual to primordial movement. It would be wrong to ascribe this stylistic progression to gradual development of skill in stone carving. The late winged-head relief style would seem to require more rather than less skill than the early linear neoclassic style....... Thus, it is gratifying that the evolutionary theory seems to work so well in this context. As I have remarked, a selection pressure has to be long lasting but not necessarily intense in order to work its ways.
[In order to examine whether art styles cross media-form boundaries we designed an] experiment that dealt with examples of baroque, neoclassic, and romantic styles, the three most commonly used cross-media stylistic terms..... Subjects could not correctly identify the artists or the titles of any of the paintings, poetry, or music............ Statistical analysis showed that subjects were grouping together instances of the same style from different media with accuracy far beyond chance. Fifty-five percent of baroque art works, 60 percent of neoclassical art works, and 65 percent of romantic art works were correctly classified. We would expect only 33 percent in each case....... The results of the experiment showed that artistically naïve adults are sensitive to cross-media styles. Presumably, the basis for this sensitivity is an untrained "direct" perception of cross-media similarities in the works of art.
To investigate the ability of children to discriminate cross-media artistic styles, we undertook a partial replication of our experiment with nursery school children having an average age of four years...... They correctly classified 8o percent of baroque, 40 percent of neoclassic, and 60 percent of romantic artworks. The children's sorting was more influenced by medium than was the adults', and the children used one rather than two dimensions to discriminate the styles.
I recall visiting an exhibition on Julio Gonzales in Madrid, about 1978 and it said something about his influence on Picasso and vice versa...in the 1920’s and 1930’s. And as I’ve recently visited the Picasso museum in Barcelona and seen some of Picasso’s very early (realist) paintings, I was struck by the apparent evolution of Picasso’s work. And, sure enough Martindale finds the same thing in his equations] ......An analysis of variance of scores grouped into five-year periods showed significant inter-period differences in both arousal potential and primordial content. The trend in arousal potential is best accounted for by a simple linear increase-in conformity with theoretical expectations. On the other hand, variations in primordial content were not due simply to a monotonic increase, though this increase accounts for 65 percent of the variation in the means shown in figure 9.9. When this trend is removed, a significant autocorrelation at a lag of two periods accounts for 25 percent of the remaining variance. [My impression is that he is pushing his statistics a bit far here and really doesn’t have enough data points with Picasso to make these claims]. Thus, we again find microstylistic oscillations. ........The microstylistic oscillations correspond to fairly clear stylistic changes. The increase in primordial content culminating in period 4 (1910-14) corresponds to Picasso's initial cubist style. The subsequent decline corresponds to his movement toward a more realistic and modelled as opposed to fat and fragmented.
My overall take on the book. I am quite fascinated by it. I really think he has identified something important but I’m not sure that the diversity in the arts is as easily corralled as Martindale would like to claim. In general, I think he has done a pretty reasonable job in his selection and randomisation of his data but, as in the case of Picasso, I think he is often pushing the explanations very hard to fit his overall theory and the data he has accumulated.
Still quite original and really interesting. Five stars from me.
Displaying 1 of 1 review