Reasoning by analogy appears to be an old-fashioned method of trying to understand revolution, or any topic within the social sciences. Today, social scientists want to find the actual causes of events, not metaphors for them. But reasoning by analogy can often reveal important insights about a subject, and perhaps more important, how it is studied.

Reasoning by analogy was the hallmark of one of the classic books on revolution, Crane Brinton's The Anatomy of Revolution (1965), which was originally published in 1938. In it, Brinton compared the course of revolution to the course of fever. In this chapter, I will discuss the strengths and weaknesses of Brinton's theory of revolution over 60 years after it was first published. I will then examine what I think is a better metaphor for revolution: murder. I will also discuss how several of the most prominent theories about the causes of revolution can be seen as variations on the elements that make up murder: motive, means, and opportunity.

Brinton published the first edition of The Anatomy of Revolution in 1938, the second edition in 1952, and the third edition in 1965. Later theorists of revolution, however, have tended to distance themselves from, dismiss, or simply ignore Brinton. In her highly influential States and Social Revolutions, Theda Skocpol described Brinton's method as "natural history" and her own as "comparative historical analysis." She noted that

At first glance, comparative historical analysis may not seem so very different from the approach of the "natural historians" ...They, too, analyzed and compared a few historical cases in depth. Actually, however, comparative-historical and natural-history approaches to revolutions differ both in objective and in method of analysis. Whereas the goal of comparative historical analysis is to establish causes of revolutions, the natural historians sought to describe the characteristic cycle, or sequence of stages, that should typically occur in the processes of revolutions. (1979, 37)

In other words, while Skocpol sought to explain why revolution occurred, Brinton merely described how it did so. Ted Robert Gurr wrote that a "fundamental limitation" of most "older" theories of revolution such as Brinton's "was the difficulty of deriving falsifiable hypotheses from them" (1970, 17-18). In his much-quoted division of writers on revolution into three generations, Jack Goldstone placed Brinton in the first generation, whose work was characterized as inductive, descriptive, and taxonomic -- which was, by implication, inferior to the historical but analytical and interdisciplinary approach of the third generation (1980). Finally, while Barrington Moore, in his monumental Social Origins of Dictatorship and Democracy (1966), examined the same four countries (plus others) that Brinton discussed in The Anatomy of Revolution, Brinton does not even appear in Moore's bibliography or index.

Yet despite what later scholars wrote (or did not write) about it, The Anatomy of Revolution is still in print and used as a text over 60 years after the first edition was published. And while many of the leading theorists of revolution saw little utility in the book, some leading policymakers assessed it far more positively. Former National Security Adviser Zbigniew Brzezinski cited Brinton as the authority for his view that the Iranian Revolution of 1979 was not inevitable and that "an established leadership... could disarm the opposition through a timely combination of repression and concession" (1983, 355). Gary Sick, a member of the National Security Council staff at the time, described the Iranian Revolution as "an almost textbook case" of Brinton's depiction of how a revolution unfolds (1986, 187). Even some of those scholars who dismissed Brinton as "descriptive" then went on to cite him in support of various arguments they made (Gurr 1970, 114, 118, 146, 150, 314; Colburn 1994, 49-50, 96).

Thus, despite the scholarly criticisms leveled against it, The Anatomy of Revolution continues to be widely read and cited -- even by scholars. What appears to account for this is Brinton's readily comprehensible metaphor for understanding revolution. According to Brinton, the course of a revolution was similar to the course of a fever. The initial stages of a revolution, like fever, were hard to detect. Once the symptoms were visible, it was too late to prevent the fever from developing. For Brinton, though, the downfall of the old regime was not the most feverish stage of the illness. For it was the moderate revolutionaries who first took power. The problems they faced, however, would prove too much for them, and soon the moderate revolutionaries would be discredited and replaced in power by extremists. These extremists would then launch a reign of terror aimed at altering the very nature of mankind. This stage would mark the highest point of the fever. Inevitably, though, the extremists would turn against one another while the population as a whole would lose faith in them. The revolutionary fever would break with the downfall of the extremists and the rise of more moderate rulers who would usher in Thermidor -- the period of convalescence and retreat from revolutionary excess. Relapses of revolutionary fever might break out again, but they too would be followed by convalescence and a return to normality. In the end, Brinton suggested, a nation experiencing revolution -- like a person experiencing fever -- returned to much the same state that they were in beforehand.

What is noteworthy about The Anatomy of Revolution is that even though it discusses only four specific cases (the seventeenth-century English, the eighteenth-century American and French, and the early-twentieth-century Russian revolutions), the stages of revolution Brinton outlined appear to be present in many other revolutions as well. Brinton stated that "the revolution in Russia has essentially run its course" (1965, 233), decades before the USSR actually collapsed. And as Sick noted, the unfolding of the Iranian Revolution occurred in a sequence remarkably similar to the stages of revolution described by Brinton.

Thus, while describing how revolution occurs may be far more modest a task then explaining why it occurs, Brinton's continued popularity results from his having performed this task so well. Indeed, Brinton's 1938 depiction of how revolutions occur appears to be far more relevant than many more recent explanations of why they do for contemporary, post-Cold War revolutions.

Nevertheless, there are serious problems with Brinton's fever metaphor. While his depiction of how revolution unfolds applies to many cases, it certainly does not apply to all. Brinton implies that once the symptoms of disease are present, the fever of revolution will follow. But as Timothy Lomperis demonstrates, some revolutions have been halted even after spreading throughout much of a country (1996). When a revolution does succeed in ousting the ancien régime, a reign of terror does not inevitably follow. There was, for example, no reign of terror in one of the four cases Brinton examined -- the American Revolution. As Knutsen and Bailey pointed out, "Brinton's problems with the American case become obvious towards the middle of the book, when the American Revolution begins to disappear from the discussion" (1989, 429n10). Nor was there a reign of terror following the 1989 democratic revolutions in most of Eastern Europe, or in democratic revolutions generally.

In those cases where there was a reign of terror, a Thermidor following it can usually be identified. Brinton, however, implies that Thermidor is the end of the revolution, marking the beginning of the return to normalcy. But even in the French case -- which gave us the term "Thermidor" -- things hardly returned to normal, as the rise of Napoleon soon followed. In the Russian case, the New Economic Policy (NEP) period (1922-28) can be considered a Thermidor that followed the terror of the civil war, but NEP was followed by yet another reign of terror under Stalin during the first two five-years plans.

Brinton's metaphor for revolution, then, is too rigid to be applicable to all cases. Indeed, following his own analogy, there are many kinds of diseases. Some are more serious than others. And even those that are serious do not all share the same pathology. Thus, while Brinton provided a useful guide for understanding how a certain type of revolution unfolds, he did not explain how other types unfold -- or fail to unfold.

Whatever its flaws, though, coming up with a better metaphor than Brinton's is not an easy task, since such a metaphor must be broad enough to account for the variation in revolutions that Brinton's does not. Is there a better metaphor for revolution than fever? In my view, there is: murder.

There are two ways in which murder and revolution are identical. Just as with those who attempt murder, those who attempt revolution do not always succeed. In addition, just as a successful murderer kills a person, successful revolutionaries destroy a government -- or perhaps more accurately, its leadership. Murder and revolution, of course, are not completely identical. A truly successful murderer is able to conceal that he or she killed someone, and has a strong incentive for not revealing what transpired. By contrast, some revolutionaries try to conceal their intention of toppling the ancien régime before actually doing so, but many do not. And once the ancien régime is toppled, revolutionaries -- unlike most murderers -- immediately claim credit for their actions. This is because, unlike murderers, revolutionaries who topple ancien régimes and set themselves up as nouveaux régimes are not likely to be brought to trial. There is, indeed, some irony in comparing revolution to murder, since revolutionaries who fail are likely to be treated as murderers, while revolutionaries who succeed are not.

After "doing the deed," however, all successful revolutionaries do something that many murderers who get caught also do: they claim that they acted out of self-defense, and that their actions should be considered justifiable homicide. Making this argument convincingly is important both for the murderer who has gotten caught and for the revolutionaries who have toppled an ancien régime. The murderer who has gotten caught must make this argument convincingly in order to be acquitted. Successful revolutionaries must do so in order to project the idea that while their overthrow of the ancien régime was justified, any attempt to similarly overthrow their nouveau régime would not be.

The most important similarity between murder and revolution, though, is that in order for either to occur, motive, means, and opportunity must all be present. It is with regard to these three elements in particular that murder is superior to fever as a metaphor for revolution. For while Brinton's fever metaphor implied that there was only one way in which revolution could occur, the murder metaphor has no such implication. Just as there is wide variation in the motives, means, and opportunities that can lead to murder (or attempted murder), there is also wide variation in how these three elements can lead to revolution (or attempted revolution).

Looking at means alone, it is obvious that murder can occur in many possible ways, including shooting, stabbing, strangling, poisoning, smothering, or drowning. Similarly, as I have argued elsewhere, revolution can occur through several means, including rural revolution, urban revolution, coup d'état, revolution "from above," revolution "from without," or a combination of these (Katz 1997, 4-9).

Further, just as all murderers do not share the same motives for killing someone, revolutionaries do not share the same motives for overthrowing an ancien régime. There are many possible motives for murder, including anger, greed, fear, revenge, or even boredom. Similarly, there are many possible motives for revolution, including not just the desire to get rid of whatever ancien régime is in power, but also out of very different visions as to what the nouveau régime should be. Indeed, since revolution is by necessity a group activity, discord can quickly develop among revolutionaries after the ancien régime has been overthrown if they do not share the same vision of the nature of the nouveau régime.

Finally, opportunity can vary considerably from case to case in revolution as well as in murder. In some cases, the would-be perpetrators plan meticulously to create the most favorable opportunity for murder or revolution. In others, however, circumstances outside of their control arise that create the opportunity for murder or revolution that its perpetrators seize upon. And in still other cases, the would-be perpetrators attempt unsuccessfully for years to create the opportunity for murder or revolution, but then unforeseen circumstances arise that make it possible after all.

As this discussion indicates, a wide variety of means, motives, and opportunities can, in different combinations, lead to revolution as well as to murder. And the examples mentioned here by no means exhaust either the possibilities of these three elements or the combinations in which they might occur.

In addition to being useful for thinking about revolution, the murder metaphor is also useful for examining and categorizing theories about and accounts of revolution. Basically, these theories and accounts can be viewed as statements about one or more of the three elements that comprise revolution.

The individual case study -- the most basic form of analysis, without which comparative histories and theories would not be possible -- examines the means, motives, and opportunities that lead to revolution in a single country. Such studies usually make no attempt to derive any sort of theory or lesson from their particular case that might be applicable to other revolutions. Indeed, the authors of such studies -- especially if they are historians -- may reject the very notion that theories of revolution apply to their particular revolution, or that their case study can be used to support a more broadly applicable theory of revolution. Just as a murder investigation seeks to explain a single murder and not murder in general, the individual case study seeks to explain a single revolution and not revolution in general. This by itself, it must be emphasized, is no mean feat: determining conclusively the causes either of a single murder or a single revolution can be extremely difficult, if not impossible. And, of course, different investigators may come up with different explanations.

While individual case studies examine the motives, means, and opportunities for a particular revolution, theories that attempt to make a statement about revolution in general based on an examination of several cases tend either to focus on one of these three variables or to see them interacting in a particular way. This can be seen through examining several prominent theories of revolution through the murder metaphor.

In Brinton's The Anatomy of Revolution, the victim (ancien régime) is an elderly family tyrant who is no longer in full possession of his faculties, while the perpetrators (revolutionaries) are two adult children still living at home who chafe at how the tyrant rules their lives and who are impatient to inherit the family house and property. The opportunity for murder (revolution) arises when the tyrant becomes so enfeebled that he can no longer defend himself, and the two children band together and kill him. At first, the two children are euphoric. But they have very different characters. One is a rational opportunist (Brinton's moderates) who simply wants to live in peace. The other is a psychopath (Brinton's extremists) who now wants to kill the neighbors and take over their houses (spread revolution abroad). The rational opportunist takes over the management of the house (the moderates take power) at first. But he or she quickly encounters several problems: (1) it turns out that the old man wasn't lying when he told them he had no money to give them; the family (the state) really is bankrupt; (2) the neighbors are horrified at what the two children did to their father, and are fearful about what they might do to them; and (3) the rational opportunist does not realize that his increasingly demanding sibling really is a psychopath until the psychopath overpowers him (the extremists overpower the moderates). The psychopath then makes a complete shambles out of running the household (the reign of terror) and becomes embroiled in bitter disputes with the neighbors, uniting them against him. The psychopath's own actions, however, trip him up, thus allowing the rational opportunist to overpower him and reassert control over the household (Thermidor). The rational opportunist then strikes a bargain with the neighbors: if they won't try to kick him out of his house, he won't try to kick them out of theirs, and so life begins to return to normal.

Brinton's theory of revolution, then, focuses on the motives of the revolutionaries and how they react to what he sees as a predictable, uniform series of opportunities they encounter once a revolution begins. And in each case, the outcome is the same: it is the moderate revolutionaries (the rational opportunists here) who eventually prevail.

Barrington Moore, by contrast, sought to explain the variation in the outcomes of revolutions: why some revolutions led to democracy, others to conservative reaction/fascism, and still others to communism. Had he used the murder metaphor, Moore's answer to this question would be that it all depends on who is the murderer.

There are three traditional farms (countries), all of which are beginning the process of mechanization (economic modernization). All three are owned and operated by tyrannical but relatively uneducated patriarchs (monarchs). Each of these patriarchs has two sons and employs a number of farmhands. Traditionally, it is the elder son (the aristocracy) who inherits, and thus the younger son (the rising bourgeoisie) can expect to receive nothing when the patriarch dies no matter how much effort he devotes to the farm.

On the first farm, the younger son (the bourgeoisie) is energetic and capable, while the elder son (the aristocracy) is not. Increasingly frustrated with life under his father's rule and the expectation of little better when his elder brother takes over, the younger son kills both his father and his brother, takes over the farm, and manages it highly successfully in his preferred mode (democratic capitalism).

What happens on the first farm has a profound effect on the second and third farms. The fathers and the elder sons take strenuous efforts to make sure that the younger sons on these two farms do not do to them what the younger son did to their counterparts on the first farm. It turns out that the younger sons on the second and third farms aren't all that energetic anyway. The elder son on the second farm, though, is. Fearing that his younger brother will eventually become more capable, the elder son kills his father, puts his younger brother under close watch, proceeds to rapidly modernize the farm, and maintains the principle of primogeniture (conservative reaction).

After what happened on both the first and second farms, the father on the third fears both his elder and his younger sons. He then keeps both under close watch, seeing any sign of initiative from either as a negative omen. Neither of these sons is particularly energetic anyway. Conditions on the farm steadily deteriorate. Eventually, it is the farmhands (the peasants) who rise up, kill the father and both sons, and take over the farm for themselves (communism).

Whatever its outcome, Moore saw the opportunity for revolution in all the cases he examined as the inability of traditional governments to effectively modernize their countries. In each case, the motive of the strongest class (bourgeoisie, aristocracy, or peasantry) was not only to overthrow the ancien régime, but to create a nouveau régime through which it could dominate the nation through suppressing (if not eliminating) the other two classes. In addition, the type of revolution that occurs in earlier cases affects and changes the type of revolution that occurs in later cases, which in turn affects and changes the type of revolution that occurs in still later cases.

Ted Robert Gurr's theory of revolution focuses on one factor: motive. Gurr sought to explain why revolution occurred not in the poorest nations, but in those undergoing modernization where significant economic development has already occurred. According to Gurr, the citizens of the former have extremely low expectations of their government and thus do not seek to overthrow it. The expectations of citizens in the latter, by contrast, grow faster than the government can satisfy them, creating a sense of relative deprivation that can lead to revolution.

Gurr's theory can also be illustrated via the murder metaphor. In this case, there are two very poor households, both headed by relatively authoritarian personalities in their forties who are in good health. In both, the adult children are, like the parents (who married quite young), relatively uneducated. They work at low-paying jobs -- and contribute a substantial portion of their incomes to the parents. They do not question this arrangement: this is what the parents had to do when they were young adults, and those who are currently young adults expect to receive this sort of support from their children in the future.

In the first household, there is no major change. It never occurs to any of the children that they would be better off without their parents. In the second household, however, the parents win the lottery or receive some other financial windfall. The adult children see this money as an opportunity to pay for college and hence improve their lives dramatically. The parents, though, have other ideas: they want to use the money for buying more lottery tickets, investing in a dubious business investment that one of their "friends" proposes, going on vacation, and other unproductive expenditures. Further, being authoritarian, they expect their adult children to continue to turn over most of their wages to them. As the parents are still relatively young and vigorous, their children -- realizing how the windfall could benefit them -- quickly acquire a very strong motive to murder their parents before the latter can spend it all. The children, of course, would not have acquired this motive if the parents had not received the windfall.

For Gurr, the coming into being of the motive for revolution (murder) will not necessarily lead to its occurrence. If the government successfully democratizes (if the parents share the wealth), harmony can be restored and revolution (murder) need not occur. But if democratization does not occur (if the parents don't share the wealth), then revolution (murder) is highly likely. Gurr does not focus so much on the opportunity factor. His theory indicates that whoever feels the motive for revolution (murder) intensely will take advantage of, or even create, the opportunity to perpetrate it.

Theda Skocpol criticized Gurr's theory of revolution for focusing on the motive for it, since this was something she saw as always being present: "What society... lacks widespread relative deprivation of one sort or another?" (1979, 34). By contrast, her theory of revolution focused on opportunity. For Skocpol, the opportunity for revolution arose as a result of state breakdown, which occurred only in very specific circumstances. Skocpol drew a distinction between the ruling class and the state. While those governing the state were drawn from the ruling class, the interests of the ruling class and the state were not always the same. It was when their interests diverged sharply that state breakdown occurred, thus allowing the oppressed classes to rise up and overthrow them.

Let us return to the household that won the lottery discussed under Gurr. While Skocpol saw the adult children possessing the motive to murder their parents, she saw them as being unable to act on this motive if the parents were of the same mind. The opportunity for murder (revolution) would arise, however, if the parents became seriously divided. If we (arbitrarily) designate the father as the state and the mother as the ruling class, sharp disagreement between the two over, say, how to spend the money from their windfall, would give the adult children the opportunity to play the two parents off against each other, kill the father -- perhaps even with the mother's connivance -- and then marginalize the mother (whose health has seriously deteriorated as a result of this crisis) afterward. For Skocpol, this scenario need not occur just in the household that received the windfall; it could occur in the one that did not if the parents there also became seriously divided.

In Jack Goldstone's theory of revolution (1991), the opportunity for revolution (in the early modern era) was created by an exogenous variable -- rapid population growth. The motive for revolution arose if the ancien régime was unable to raise sufficient revenue to cover the rapidly rising expenses it had to make on this growing population (especially the growing elite population). By contrast, during periods when population growth was low, the ancien régime was able manage its expenses and receipts more successfully, and so maintain stability.

Goldstone's theory can also be illustrated with the murder metaphor. In this case, there is a business owned and managed by a large extended family. The older generation that built up the firm had many children. As these come of age, they all demand positions in the firm. The problem is that the business is not profitable enough to employ all of the members of this large younger generation at the high salaries they expect. While the senior partners all recognize this problem, each of them is determined to bring his own children into the firm. But the more each of them succeeds in doing this, the less profitable the firm becomes. And the less profitable it becomes, the more the head of the firm is blamed for this situation by others in the firm.

In Goldstone's model, however, it is not just the head of the firm whom others in it want to get rid of. In fact, virtually everyone has an incentive to murder everyone else. There is generational conflict between the senior partners and the junior ones who want to replace them. And there is also conflict within each generation, especially the large junior generation vying for position and promotion. As this situation worsens, everyone pursues his or her own individual interest at the expense of the firm's as a whole, which as a result is becoming less competitive than other firms (other countries). If the senior partners can somehow manage to increase revenue, catastrophe can be avoided. But if not, a more vigorous segment of the junior generation may kill off the senior partners (a necessary move since they won't surrender ownership voluntarily) and devise a bolder business plan that can fully absorb the large number of junior partners.

If this junior, or second, generation of partners has few children (for whatever reason), these pressures will not arise as they come of age. This relatively small third generation will easily be absorbed into the firm and will rise up rapidly within it as the second generation retires. But if the third generation has many offspring, the large fourth generation will create problems for the third generation similar to those that the second generation created for the first.

In Goldstone's model, then, high population growth created the opportunity for revolution in the early modern era, but did not make it inevitable. If the ancien régime could successfully manage the financial stresses caused by this growing population, it could avoid revolution. But if it could not manage them (and it sometimes could not), revolution would result.