ONLINE FIRST – Clint Ballinger « Contingency is just so »

COUV T24En prévision de la sortie (prévue pour le 21 mai 2013) du numéro 24 de Tracés, intitulé Réalité(s) du possible en sciences humaines et sociales, nous proposons en avant-première la version originale de l’article de Clint Ballinger, soumis en anglais et publié dans ce numéro en traduction française sous le titre « Contingence, déterminisme et ‘just-so stories‘ » (Tracés, n°24, p. 47-69).

Anticipating the release (scheduled for May 21 2013) of the 24th issue of Tracés, entitled Réalité(s) du possible en sciences humaines et sociales, we propose online first the original version of Clint Ballinger’s paper, submitted in English and published in French translation in this forthcoming issue under the title « Contingence, déterminisme et ‘just-so stories' » (Tracés, n°24, p. 47-69).

*Pour citer la version anglaise / to cite the English version : Clint Ballinger, “Contingency is just so”, Le carnet de Tracés, publié le 29 avril 2013, Url:


Contingency is just so

Clint Ballinger

B.A., Archaeology, University of Texas at Austin; M.A., Political Science, University of North Carolina at Chapel Hill; Ph.D., Geography, University of Cambridge

A number of contrasting uses of contingency are discussed, with an emphasis on the relation of contingency with indeterminism and determinism. We find associations with indeterminism problematic. Another frequent use of the concept of contingency is discussed, what we call epistemic contingency. It is associated with spatiotemporal conjunctures and is intricately connected with the concept of boundary conditions (how humans perceive the world) and initial conditions. The latter are the often ignored half of the (laws and initial conditions) causal story and account for the spatiotemporal facts about the universe. We explain a view that what humans perceive as “contingent” stems from the irregular initial conditions of the universe. The story of how is crucial to any meaningful understanding of the world and is, contrary to common views, both increasingly tractable and correctly told as a just-so story.

Keywords: contingency, counterfactual possible worlds, explanation, indeterminism, initial and boundary conditions

I. Introduction

The concept of contingency is widely invoked in the human sciences,1 sometimes even claimed to be central to their understanding. Sociologist Wagner-Pacifici, for example, asks: “Might not one say that the whole project of sociology is to account theoretically for the contingent patterns and shapes of this mutable and mutating social stuff of life…? (Wagner-Pacifici, 2000, p.1). In economic geography, Plummer states that “In order to make any headway in empirical modelling…we need to formulate theories in which the nature of contingency and complexity…are specified more clearly” (Plummer, 2001, p. 764). For the historical sciences more generally Stephen Jay Gould states that “the science of contingency must ultimately be integrated with the more conventional science of general theory” (Gould, 2002, p. 46).

These views on contingency in the study of society, a concept already given importance by historians such as J.B. Bury and W.B. Gallie, seem to have been on the rise in recent decades. Contingency has perhaps been still more important in recent decades in biology, associated with Gould’s emphasis on the contingency of evolution. In turn, the emphasis on contingency in biology has had a substantial impact on all of the historical sciences, including the social sciences (see Blaser, 1999; there are countless references to Gould’s emphasis on contingency in the social sciences2).

However, there seems to be significant confusion about the meaning and significance of assumptions surrounding contingency. Particular assumptions underlying concepts of contingency have important implications for the human sciences, yet the concept is used in ways that are inconsistent, unrigorous, and even contradictory.

Contingency is often invoked when questions concerning causation, explanation, momentous change, and similarly weighty issues are at stake.3 It often seems to be closely connected to underlying stances on indeterminism, which in turn potentially influence one’s approaches to such things as causation, interpretation of statistics, the application of modal logic to the social sciences,4 ethics,5 psychology, evolution, theories of the mind, the nature-nurture debate, structure and agency, views on geographic influence on society, and the idiographic-nomothetic Methodenstreit. It is important to be explicit about underlying assumptions of indeterminism when contingency is discussed or the word ends up a tool for elision rather than insight.

Regarding the scope for confusion and multiplicity of meanings of contingency, consider the following eight sentences:

  1. “Globalization’s effects on democratization are thus indirect, contingent upon the scope and desirability of redistributive policies.” (Rudra, 2005, p. 708)

  2. “Contingency refers to the inability of theory to predict or explain, either deterministically or probabilistically, the occurrence of a specific outcome” (Mahoney 2000, 513)

  3. Contingency in history is “[t]he view that chance and accident play as important a role as structure” (Brucker, 2001, p. 1)

  4. “Oakeshott argues that agents are “free” by virtue of their ability to choose particular contingent responses to their situations.” (Mapel, 1990, p. 393)

  5. “individuals or other phenomena have both necessary qualities, existing across all possible worlds, and contingent qualities, existing in only some possible worlds” (Sylvan and Majeski, 1998, p. 82)

  6. “In this analysis ‘contingency’ will be taken to mean a conjuncture of events without perceptible design” (Shermer, 1995, p. 70)

  7. “path dependency characterizes historical sequences in which contingent events set institutional patterns with deterministic properties in motion.” (Djelic and Quack, 2007, p. 161-162)

  8. “The philosophical significance of chaos theory is that it reconciles the notions of causation and contingency” (Ferguson, 1997, p. 79)

Each of these uses the word contingency, but in distinct ways. (1) uses the word in the popular usage manner to mean something like “depends on” while (2) emphasizes the common association of contingency with unpredictability. (3) emphasizes the role of chance in contingency and (4) relates contingency to free will. (5) places contingency in a modal logic possible worlds framework. (6) uses the concept of conjuncture to define contingency. (7) associates contingency with the concept of path dependency and (8) with chaos theory.

These examples illustrate the scope for confusion surrounding the word “contingency”. This is doubly so given the fact that concepts such as possible worlds, indeterminism, and path dependency are themselves difficult and disputed. More confusing still, the different uses of the word “contingency” are sometimes directly contradictory.6 These facts alone should make any writer concerned with precision wary of the term. With so many different interpretations of contingency its use is likely to cause further confusion rather than increase the clarity of an argument.

We view the elision of stances on indeterminism and determinism7 as perhaps the most serious problem with usage of the word “contingency”, and therefore focus on this division and usages (3) – (6). We consider path dependency and chaos theory in an appendix to make the main argument easier to follow. (1) is not especially problematic,8 but included to show the numerous uses of the word contingency. After briefly discussing (2) we move on to the division between indeterministic contingency and a use that is compatible with determinism, epistemic contingency. We conclude with an outline of why we see contingency as “just so”, and the central importance of this for explanation in the human sciences.

Unpredictability (sentence 2) is a general result or attribute of contingency, rather than a definition of the word itself. Saying an outcome was “contingent” and thus unpredictable serves as a means of elision of a central question: Is the unpredictability due to ontological indeterminism or is it merely epistemic?

II. (Ontological) Indeterminism and Contingency

Ontological Chance and Free Will

As pointed out in the introduction, assumptions of indeterminism potentially have deep significance for interpretations of key problems in the social sciences. Where use of the word “contingency”” masks these assumptions it is less easy to understand what is at stake in a debate (for example, for incompatibilists the issue of free will is at stake, which may in turn affect interpretations of political ethics or theories of the mind).

The significance of contingency and its association with chance or free will can be seen in the examples by Brucker and Mapel where contingency and its associated indeterminism plays a role “as important as structure” and makes agents “free”. We will return to the issue of contingency and indeterminism in the conclusion, after discussing a concept of contingency consistent with determinism in Part III.

Counterfactual possible worlds

(5) is an example of a possible worlds use of the word contingent in the social sciences. Social scientists have sometimes used a modal logic definition of contingency (e.g., “Contingent means ‘neither necessary nor impossible’”, Sayer, 2000, p. 123) and/or put discussions of causation in terms of counterfactual possible worlds. Perhaps this is because they have been “decidedly in philosophical vogue” (Rescher 1999, 403), or in an attempt at a novel approach to the social sciences, or because of the commendable idea to look to philosophy for definitions.

Social scientists consider alternative histories of our actual world with identical initial conditions and laws of nature (these can be notated as a branching conception of the actual world, or as different possible worlds with the same histories up to some point). In metaphysics, alternative possible worlds are also posited that have different initial conditions and laws of physics, but we doubt social scientists have these in mind. If social scientists are concerned with worlds that are universes like ours, which we think they are, then the concept of tiny changes to laws/initial conditions is of little utility. Any universe with even the slightest difference in its laws of nature would be radically different than ours (Rees 1999, see also Bird 2002); any universe with the slightest (i.e., even quantum level) difference in its initial conditions would be radically different in its spacetime structure than ours (see the discussion of Agullo et. al., forthcoming, in Penn State 2012; this point is explored at length in Ballinger 2008a and 2008b). It is doubtful they could even be claimed to be “sufficiently” like ours if “sufficiently” is defined to mean able to provide the kind of causal insight into the actual world we imagine most social scientists desire. Social scientists are thus most likely imagining possible worlds with the same laws of nature and initial conditions as our actual world.

This, in turn, means social scientists are, perhaps unwittingly, again obscuring an assumption of indeterminism. Why? Because by definition (see footnote 7 above), if considering possible worlds with the same laws of nature and initial conditions as ours, for there to be anything different between them entails indeterminism. That is: Under determinism, any world which has the same laws of nature as w and which is exactly like w at t0 is exactly like w at all times which are future relative to t0.9 (To be clear, we are not saying that in general possible worlds semantics implies indeterminism. We are pointing out that indeterminism must be assumed by definition for there to be differences between worlds specifically in cases where a possible world is imagined that has the same laws/initial conditions as ours (as most social scientists are likely imagining).10

More importantly, in addition to what we view as problematically obscuring an assumption of indeterminism, we think the counterfactual worlds framework is much less useful than some social scientists seem to believe. Why?

Social scientists have usually turned to counterfactual possible worlds in an effort to clarify causal notions about the world we live in.11 However, counterfactual possible worlds are only useful to illustrate already held causal notions from the actual world. If we imagine a world identical to ours save for the counterfactual that Napoleon died young, or “John’s car does not break down because he remembered to put oil into the engine”, we make clear what our causal beliefs are concerning the effect of Napoleon on European history or of oil on car engines. Counterfactuals of this sort are also used because they seem able to deal with specific events (Napoleon and Europe, John’s car) and thus seem especially useful for thinking about singular, or token, causation and historical explanation (in evolution, astronomy, geology, or history).

Possible worlds are indeed useful for illustrating causal ideas we have from the actual world (Napoleon changed Europe dramatically, oil helps engines run and so on). Counterfactual contingent (“neither necessary nor impossible”) propositions in possible worlds express causal beliefs because of what we know about the actual world, and can illustrate such causal beliefs clearly at times. Crucially, though, what counterfactuals can not do is give us new information about the actual world upon which to form novel causal beliefs.12

To summarize, 1. Counterfactual possible worlds fail to add any causal knowledge about the actual world. Counterfactuals can only illustrate causal notions already held from the actual world.

2. The specific way in which social scientists use the concept of contingency with counterfactual possible worlds hinges, like the uses in sentences (3) (ontological chance) and (4) free will (at least for incompatibilists), on indeterminism, and thus still entails all the problems associated with indeterminism for causality and explanation.

There seems to be little to gain and much to lose for social scientists in turning to counterfactual possible worlds for explanation (viz. scope for confusion with arcane notation and terms, investment in time, potential for misunderstanding, and the obscuring of assumptions concerning indeterminism).

Perhaps, however, something other than indeterminism entirely is meant by contingency. Perhaps contingency is not about what we can’t know, but what we don’t know.

III. Epistemic contingency

The last section discussed conceptions of contingency arising from indeterminism. What about contingency under determinism? It might seem difficult for contingency to mean anything under determinism as there are no possible alternative outcomes. However there are at least two ways, both related to our knowledge about the world, in which the concept of contingency can apply. The first involves hidden factors.

Hidden factors

It could be that what is described as a possible contingent outcome is not due to indeterminism arising from ontological chance or free will, but rather to hidden factors. The factors might be either too complex, too small, or both, to be seen or understood.

This could be true of any event attributed to indeterminism due to chance or free will.

A classic illustration of the debate on hidden factors is the roulette wheel. Roulette is generally seen as a good example of a game based purely on chance. The factors that determine where the ball will land are both too small and too many (precisely when the ball is released, the speed of the wheel, the tiniest irregularities in the wheel and ball). However, this is purely an epistemic problem – we could in theory discover more precisely the details of these facts and the final position of the roulette ball would not be viewed as ontologically indeterministic.

The same hidden factors argument is at the center of debates about biological determinism, just with vastly greater complexity (and in quantum mechanics, with vastly smaller size). It could be that the immense complexity and small scale of the interactions in biology and the brain make it impossible to understand the brain epistemologically, without entailing that there is ontological indeterminism (see Silberstein and McGeever1999, Millstein 2000 and 2003, Weber 2005; contrary to widespread belief, the same is true of quantum phenomena, see Earman 2004). Like the modal use of the word contingency when referring to the actual world, this brings us back to the determinism/ indeterminism debate again.

Besides knowledge/lack of knowledge of hidden factors, another similar lack of knowledge is very frequently described as “contingency”. We turn to it now.

Boundary conditions, systems, and point of view

As shown in example 6 in Part I, one use of the word contingency is to describe the result of a conjuncture. This is frequently expressed as the coming together of “separate causal paths.” These causal paths could, at least in theory (like a roulette wheel and ball), be traced unbroken – that is, there is no claim made for a break in the historical causal chains of the factors in the conjunction as there is when acausal chance, free will or divine intervention are called upon to account for a “contingent” outcome.13 A recent statement of this idea from biology expresses this view of contingency and deterministic causal chains well:

Contingency may be defined as the outcome of a particular set of concomitant effects that apply in a particular space-time situation and thus determines the outcome of a given event. In most of the epistemological literature, this word has aptly taken the place of the term `chance’ or `random event’, and in fact, it has a different texture. For example, a car accident can be seen as a chance event, but indeed it is due to the concomitance of many independent factors, like the car speed, the road conditions, the state of the tyres, the alcohol consumption of the driver, etc. These factors all sum together to give the final result, seen as a chance event. The same can be said for a stock-market crash, or the stormy weather of a particular summer day. Interestingly, each of these independent factors can actually be seen per se as a deterministic event, e.g. the bad state of the car tyres determines per se a car sliding off at a curve. The fact, however, that there are so many of these factors, and each with an unknown statistical weight, renders the complete accident unpredictable: a [non-ontological] chance event. (Luisi, 2003, p. 1142)

What frequently causes mere conjuncture to be labeled “contingency” is the point of view of the agent. For example, one of Lamprecht’s examples of contingency in “Contingency in Nature” (1971) is a hunter with excellent aim who fires at a buck. To the hunter’s surprise the buck gallops off unharmed. Puzzled, the hunter follows the path of the bullet towards the prey, and finds a freshly broken branch in the path of the bullet. Apparently a breeze had blown the branch just as the hunter fired, deflecting the path of the bullet. This “contingent” event caused the hunter’s bullet to miss the buck. In this example, at least in theory if not in practice, the causal chain of the wind and growth of the tree branch could be traced (unlike in indeterministic contingency where free will or ontological chance are acausal, representing a fundamental break with any causal history). But to the hunter who at the time of the event had no knowledge of the impending gust of wind or the existence of a wind-sensitive tree branch, the outcome seems contingent. This would not be so if the hunter had had full knowledge of the existence and impending confluence of these factors. In biology the extinction of the dinosaurs is frequently held up as prime example of a contingent outcome due to the K-T event 65 million years ago. Again, there is not necessarily ontological chance involved. From a terrestrial view the event is contingent; to aliens viewing the development of Earth but also with full knowledge of the trajectories of asteroids it would not be viewed as contingent.14

The view of “contingency” as ignorance of other causal streams or systems of the universe, and hence ignorance of how they will eventually impinge on each other, is common in the social sciences. For example, it is the view of one of the more influential social science discussions of contingency, J.B. Bury’s “Cleopatra’s Nose.” The view that contingency refers to an individual’s degree of knowledge about separate systems or causal streams is clear in his discussion of “pure” and “mixed” contingencies. Pure contingency results from seemingly completely unrelated systems interfering with one another; as we begin to deem the systems as more and more related, we think of the event as less and less contingent:

If Napoleon at an early stage in his career had been killed by a meteorite, that would have been the purest of pure contingencies…The meteorite was completely disinterested in his death…But suppose Napoleon would have been killed by the hand of an assassin who detested his policy. This would not be a pure contingency. For the assassin was interested in Napoleon’s death, and the causal sequence which led him to commit the act would have been connected with the causal sequence which rendered the great man’s death historically important. (Bury, 1916, p. 67-68).

In Bury’s example, the more separate the causal paths that come together the more contingent an agent will view the outcome. Bury, like Luisi, emphasizes that the “chance” of contingency results from the collision “of two or more independent chains of causes” (Bury, 1916, p. 61). There are numerous allusions to contingency as resulting from the conjuncture of multiple factors or systems in the social sciences. For example, Mandel characterizes “contingent…historical accidents” as “occurrences that represent an unpredictable or fortuitous confluence of factors” (2003, p. 178). Steinmetz writes that “contingency here means that complex events are codetermined by constellations of causal mechanisms” (1998, p. 177) and Mann that “Many causal paths cross, and the combination of factors in each spatiotemporal location makes it unique – this is where contingency is the same as ‘context’” (1994, 47).

Crucially, descriptions of contingency as conjuncture, as Bury and Lamprecht emphasize, stress the “externality”, separation from, or ignorance of other systems (and thus their impending interference with one another). For example Djelic et. al. write “These contingent developments or pressures for change tend to be seen as external to the system.” (2007, p. 166). Jessop discusses contingency as meaning “indeterminable within the terms of a single theoretical system” (1990, p. 12), Arthur writes that “Gallie’s use of ‘contingent’ seems to be equal roughly to ‘unpredictable from the point of view of the main process’” (1968, p. 208) and we read that a contingent event “may be seen as the result of a coincidence between two phases of mutually independent processes.” (Hypergéo).15 Humans ascribe boundary conditions to irregularities in the universe and call them systems (including “things” and “objects”- solar systems, galaxies, oceans, volcanoes, cities, engines); the interaction of these perceived systems is central to the concept of contingency.

The kinds of interactions that appear contingent and have dramatic unexpected changes involve the clash of the very large with the very small, the very long-term with the very short-term, slow and fast, complex and simple. Consider the microscopic workings of biochemistry in the vastly larger Earthly and social environment, the interaction of long-term large-scale geographic factors on more recent and rapidly changing human organizations, vast weather systems and small human events (a battle, a hunter shooting at a buck) and disease epidemics. In this conception, contingency is about the spacetime interactions of things, often with humans and their interests front and center – weather, geographic factors, a gust of wind to a hunter, imaginary meteorites hitting Napoleon (or real meteorites extinguishing the dinosaurs).

Epistemic contingency under determinism is about what we know. Either we do not know why there is one outcome over another because of the complexity of a system, because of the scale of the factors, or because something outside of the system we have imagined suddenly impacts it, altering it in a way that was unknowable to the observer.

IV. Conclusion

Contingency is frequently associated with unpredictability. This can be due to ontological indeterminism (ontological chance, or free will for incompatibilists). If something is viewed as unpredictable because of ontological indeterminism, it seems that it would be clearer to simply make this belief explicit rather than describe the event as “contingent”. Something is deemed to have happened that could have been (in an alternate world) otherwise, and further explanation of the event is not possible; the ultimate answer is that it was due to ontological indeterminism. We share the same misgivings with the explanatory dead-end of indeterminism as physicist Jean Bricmont: “how does one show that any unpredictable system is truly non-deterministic, and that the lack of predictability [or explainability] is not merely due to some limitation of our knowledge or of our abilities? We cannot infer indeterminism from ignorance alone.” (2000, p. 2)

Alternatively, contingency can result from limited knowledge and perspective. When things interact (spatiotemporal “conjunctures”) in ways that are unexpected due to a lack of knowledge the outcome is described as “contingent”. There is no appeal to indeterminism needed, just ignorance of our world. Unlike ontological indeterminism, however, under epistemic contingency we have not arrived at an explanatory dead-end, but rather at the beginning of an explanation.

Explanation, boundary and initial conditions, and just so stories

Laws of nature have received the bulk of (both good and bad) attention in disputes on explanation. However, the most basic conception of what shapes the universe is that laws of nature and initial conditions have equal but different roles (Frisch, 2004, p. 696 provides a brief exposition). Initial and boundary conditions, however, receive relatively very little attention in the philosophy of science (Ballinger, 2008b). How do these relate to contingency, and why do they help make contingency tractable (and indeed central) to better understanding our world?

Things (“objects”, “systems”, and their “exogenous factors”) can be thought of as human constructed boundary conditions referencing the spatial irregularities of our world. At times the boundaries seem natural, i.e., where the wood stops and the air begins more or less bounds the concept of “table”. This is especially so with familiar objects at human scales. But at larger scales, greater complexity, and with less obvious material boundaries (“forest”, “city”, “solar system”, “galaxy”) it is clearer that what we call a thing or object or system is in fact a human-constructed boundary condition referencing some spatial irregularity in the universe. These are in fact open systems that we “close” for convenience. The only non-constructed system or thing is the universe itself.

If “things” reference the irregularities in the universe, and contingency is ascribed, as it often is, to the conjuncture of things, then this raises a rather fundamental question: Why are there spatial irregularities at all? Where do they come from?

This question in fact has an answer: They come from the initial irregularities of the universe. Every current irregularity (to which humans ascribe boundary conditions and call things, objects, or systems) is the result of some earlier irregularity. Like quantum phenomena, or the vast time- and spatial- scale of the universe, this reality is difficult to grasp, that all the spatial irregularities we encounter in the present can be traced to the irregular initial conditions of the universe. Nevertheless, that is what the best evidence shows (see the discussion of Agullo et. al., forthcoming, in Penn State 2012; this point is explored at length in Ballinger 2008a and 2008b). Even the quantum phenomena that appear to us indeterministic decohere in spacetime interactions that are external to them, and due to the irregular initial conditions of the universe. As a closed system the universe as a whole unfolds deterministically according to Schrödinger’s equation (Weinberg 2001, 118-119; Schlosshauer 2004).

The irregular initial conditions of the universe are thus 1. Responsible for the existence of the irregularities that humans reference with boundary conditions and call “things” or “systems” and 2. Responsible for the spatiotemporal unfolding of the universe and the interactions of things (objects, systems) that, when unexpected, we call contingent.16 Present spatiotemporal events, even at small scales important to the social realm,17 are explained by irregular antecedent and ultimately initial conditions.18 Therefore if contingency is a way to describe our ignorance of spacetime events, then initial conditions are central to an understanding of contingency.

Our evolutionary psychology is not adapted to understand the vastly smaller, the vastly larger, the vastly longer-term, and the vastly shorter-term. It is still worse equipped to understand the intersection of these. The concept of contingency captures this limitation.

Integrating contingency into a general theory of science, as Gould called for, first and foremost necessitates a general theory of why things vary across space and why they have the spacetime trajectories that they have. In Ballinger 2008b (see also 2011b) we have set out the beginnings of just such a theory in an extended treatment of the idea, including applications relevant to the social sciences. It is now understood that the universe had irregular initial conditions, and that these have led to all subsequent irregularity (variegation) in the universe. The way to understand the interactions of these irregularities (systems) is to trace the real spacetime trajectories of antecedent states. There is not, at macro-scales, causation in the way it is usually thought of, only context.

Using descriptive (not inferential) statistics properly, as we have argued elsewhere (Ballinger 2011a), we can arrive at satisfying explanations of events viewed as contingent even in the human sciences. This was previously difficult, in part because a significant amount real-world data is needed, and this simply was not available in the past. However, as is much discussed these days, we have entered into the age of “Big Data” (and “digital humanities”, where big data is very much applied to the social realm, often with attention to spatial interrelations).

We can finally truly begin to study the other half of the causal equation – initial conditions instead of laws of nature – in ways that were simply not possible before. Complex sequences of past events are now tractable problems. The study of irregular initial, antecedent, and boundary conditions is not only possible, but necessary for satisfyingly complete explanations of our world. It is time for our theories of explanation to catch up to this fact.

This may seem like a Herculean (or, perhaps, Laplacian) task. But this is to misunderstand the nature of the proposal. The traditional criticism of idiographic explanations is that they entail description of (and only apply to) every particular event. A related criticism is that these explanations are “just-so stories”, and therefore not explanatory at all, nor scientific.

But the “just so” criticism is confused. Any correct explanation in the historical sciences will necessarily be a just-so story. This crucial point is far too often forgotten and consequently good explanations are wrongly criticized. The purpose of the “just-so story” warning is to make it clear that there are many possible wrong answers to any historical explanation. The need for lots of data in historical explanation is precisely to narrow down explanations of historical events (this includes historical events such as the formation of the Earth and all of evolution) to sets of possible explanations narrow enough to satisfy us. Indeed, we may even finally, with all due respect to Mr. Kipling, learn the true story of “How the leopard got his spots”.19

As Sven Steinmo so cogently argues in discussing the human sciences “if politics and history are instead the products of emergent phenomenon and contingent events, then we should teach (and learn) about the people, places and events themselves – not only because that can help us understand the processes at work, but also because in an emergent and contingent world, these facts are themselves important. (Steinmo, 2010, p. 2, emphasis in the original). Laws of nature alone are at best half the causal story; for a compelling etiology of our world and social realm we must fully and deeply integrate initial conditions into our explanations. We live in a just-so world full of events viewed as contingent. To understand it we need just-so explanations.20

Appendix – Path Dependency & Chaos

How do chaos and path dependency relate to contingency and indeterminism?

Sentences (3) and (4) from Part I closely associate or give connotations of chaos theory or path dependency to contingency, but do not actually define contingency as those things. The fact that they so frequently conflate contingency with these other terms is potentially problematic.21 Note the end result of a chaotic process, or of a path dependent process, no matter how complex or surprising the end state, would properly be explained as just so stories.

Path Dependency

Path dependency is linked to contingency primarily through the concept of multiple equilibria, which is the same as saying there are multiple possible contingent outcomes. A common way of depicting path dependency is as a Polya urn process. Imagine a large urn with a red, a blue and a yellow ball in it. A ball is taken out, then that ball is returned to the urn with an additional ball of that same color. In the first turn, the probability that a ball of any color taken out is equal. However, if a red ball is taken out and then returned with an additional red ball, on the next turn there is a higher chance of a red ball being taken out. If, as is increasingly likely, the next several turns extract a red ball, soon there is virtually no chance of drawing blue or yellow balls. The first few draws from the urn result in a “lock in” effect for either red, blue, or yellow. Thus the replacement rule simulates the concept of increasing returns leading to “lock in,” with early events leading to lasting later pathways. However, as with chaos, it is important not to conflate path dependency with contingency. Think of the Polya urn process again – what is interesting about it is the lock-in effect, which is caused by the replacement rule in the Polya urn example. The question, then, is why that replacement rule? Why are we returning two (or three or a hundred) balls for every ball withdrawn? (Just as in economic systems, we want to know why there are increasing returns). Conflating path dependency and contingency obscures what is unique about path dependency, that something is causing increasing returns and lock in effects.

Additionally, saying that path dependency is “marked by contingency” as Mahoney 2000, p. 513) and others do again serves to obscure assumptions concerning determinism and indeterminism. Imagine the Polya Urn Process again. On each draw, the fact that one or another color ball is drawn could be due either to ontological chance or to hidden factors.


Chaos – unpredictable outcomes in deterministic systems resulting from small changes in the initial conditions of the system – is conflated with or closely linked with contingency primarily through the concept of unpredictability, as shown in (3). Chaos does not have any relation to modal logic or counterfactual possible worlds. Chaotic systems are by definition deterministic and thus contrary to indeterministic usages of contingency, and compatible with epistemic contingency.

What is central to the concept of chaos is the sensitivity to initial conditions of chaotic systems (which happens to lead to unpredictability for epistemic reasons). Again we find initial conditions central to a concept of contingency.


Adcock, Robert, 2007, Who’s Afraid of Determinism? The Ambivalence of Macro-Historical Inquiry, Journal of the Philosophy of History, Vol. 1, no.3, p. 346-364.

Agullo Ivan, Ashtekar Abhay, and Nelson William. 2012, Quantum gravity extension of the inflationary scenario. Physical Review Letters, December 11.

Allen William L., Cuthill Innes, Scott-Samuel Nicholas and Baddeley Roland, 2010, Why the leopard got its spots: relating pattern development to ecology in felids. Proceedings of the Royal Society B, vol. 278 no. 1710, pp. 1373-1380. (A popular account can be found at: University of Bristol, « Why the leopard got its spots. » ScienceDaily, 20 Oct. 2010. Web. 11 Dec. 2012. )

Arthur C. J., 1968, On the historical understanding, History and Theory vol. 7, p. 203-216.

Astor Aaron, 2010, Contingency, Counterfactuals and the Study of History, The Moderate Voice, Dec 7, 2010.

Ballinger Clint. 2008a. Initial conditions and the “open systems” argument against laws of nature. Metaphysica 9(1): 17-31.

Ballinger Clint. 2008b. Initial conditions as exogenous actors in spatial explanation, Ph.D. Dissertation, Department of Geography, Univ. of Cambridge. Available online,

Ballinger Clint. 2008c. Classifying contingency in the social sciences: diachronic, synchronic, and deterministic contingency. Working paper, philpapers archive,

Ballinger, Clint, 2008d. Determinism and the Antiquated Deontology of the Social Sciences. Working paper, philpapers archive,

Ballinger Clint, 2011a, Why inferential statistics are inappropriate for development studies and how the same data can be better used. Working paper, Munich RePEc Archive

Ballinger Clint, 2011b, Why geographic factors are necessary in development studies, Working paper, Social Science Research Network SSRN:

Barker Stephen, 2011, Can Counterfactuals Really Be about Possible Worlds? Noûs Volume 45, Issue 3, p. 557–576.

Bathelt Harald and Glückler Johannes. 2003, Towards a relational economic geography, Journal of Economic Geography 3: p. 117-144.

Benaceraff Paul,1973, Mathematical Truth, Journal of Philosophy, 70: p. 661–679.

Bird Alexander, 2002, On whether some laws are necessary, Analysis 62.3, 257-270.

Blaser Kent, 1999, “The History of Nature and the Nature of History: Stephen Jay Gould on Science, Philosophy, and History”, The History Teacher 32(3): p. 411-430.

Booth Charles, Rowlinson Michael, Clark Peter, Delahaye Agnes, Procter Stephen, 2009, Scenarios and counterfactuals as modal narratives, Futures , vol. 41, no. 2, pp. 87-95.

Bricmont, Jean. (2000). “Determinism, Chaos Theory and Quantum Mechanics”. Institute of Theoretical Physics, Université Catholique de Louvain.

Brown Cynthia Stokes. 2007. Big History: From the Big Bang to the Present. New York: New Press.

Brucker Gene, 2001, “The horseshoe nail: Structure and contingency in medieval and renaissance Italy”, Renaissance Quarterly 54:1-19.

Bryant, Joseph M. 1992, Towards a Respectable, Reflexive, Scientific Sociology: A Note on the Reformation Required, Canadian Journal of Sociology, Vol. 17, No. 3. (Summer, 1992), pp. 322-331.

Bury J.B., [1916] 1930. “Cleopatra’s nose”, In Selected essays of J.B. Bury. Harold Temperley, ed., Cambridge: Cambridge University Press.

Chaisson Eric. 2001. Cosmic Evolution: the Rise of Complexity in Nature. Cambridge, Mass: Harvard University Press.

Christian David. 2004. Maps of Time: An Introduction to Big History. Berkeley, University of California Press.

Climo T. A. and Howells P. G. A., 1976, “Possible worlds in historical explanation” History & Theory 15: p. 1-20.

Djelic Marie-Laure and Quack Sigrid, 2007, “Overcoming path dependency: path generation in open systems”, Theory & Society, 36: p. 161–186.

Earman, John. 2004. ‘Determinism: What We Have Learned and What We Still Don’t Know’, pp. 21-46 in Joseph Keim Campbell, Michael O’Rourke, and David Shier (eds.), Freedom and Determinism. Cambridge, Mass: MIT Press.

Elster Jon, 1978, Logic and society: Contradictions and possible worlds. Chichester: Wiley.

Ferguson Niall, 1997, “Virtual history: Towards a ‘chaotic’ theory of the past”, In Virtual history: Alternatives & counterfactuals, Niall Ferguson ed. London: Picador.

Frisch, Mathias. 2004. Laws and Initial Conditions. Philosophy of Science, 71: 696-706.

Garud Raghu, Kumaraswamy Arun and Karnøe Peter, 2010, “Path Dependence or Path Creation?” Journal of Management Studies 47:4 June 2010.

Gottschall Jonathan, 2012, “Defending Just-So Stories: Why science needs stories”, The Storytelling Animal (PsychologyToday blog) September 24th, 2012.

Gould Stephen, 2002, The structure of evolutionary theory. Cambridge, Mass.: Harvard University Press.

Griffin Larry, 1993, “Narrative, Event-Structure Analysis, and Causal Interpretation in Historical Sociology”, The American Journal of Sociology 98: p. 1094-1133.

Hawthorn, Geoffrey. 1991 (1995). Plausible worlds: Possibility and understanding in history and the social sciences. Cambridge: Cambridge University Press.

Hoefer Carl, « Causal Determinism », The Stanford Encyclopedia of Philosophy (Summer 2005 Edition), Edward N. Zalta (ed.),

Hypergéo, Society and environment – epistemological foundations, online encyclopedia of geography. ‘Contingency’, 2005, Accessed November 3, 2012.

JESSOP Bob, 1990, State Theory: Putting the Capitalist State in its Place, University Park: Pennsylvania State University Press.

Lamprecht Sterling, 1971, “Contingency in nature”, Philosophy and Phenomenological Research 32: p. 1-14.

Lewis David, 1973, Counterfactuals, Blackwell, Oxford.

Liebowitz S. & Margolis Stephen, 1995, “Policy and path dependence: From QWERTY to Windows 95”, Regulation: The Cato Review of Business and Government 18(3): p. 33-41.

Luisi Pier Luigi, 2003, “Contingency and Determinism”, Philosophical Transactions of the Royal Society of London, Series A: Mathematical, Physical and Engineering Sciences 361(1807): p. 1141-1147.

Mahoney James, 2000, “Path dependence in historical sociology”, Theory and Society 29: p. 507-548.

Mandel D.R. 2003, “Simulating history: The problem of contingency”, Analyses of Social Issues and Public Policy 3: p. 177-180.

Mann, Michael. 1994. In Praise of Macro-Sociology: A Reply to Goldthorpe. The British Journal of Sociology 45(1): 37-54.

Mapel David, 1990, “Civil association and the idea of contingency”, Political Theory 18: p. 392-410.

McNeill J.R. 2005. ‘Global History: Research and Teaching in the 21st Century’. Occasional Paper, Center for Global Studies, University of Illinois at Urbana-Champaign, June 20.

Millstein Roberta 2000, “Is the evolutionary process deterministic or indeterministic? An argument for agnosticism”, Biennial Meeting of the Philosophy of Science Association, Vancouver, Canada.

Millstein Roberta 2003, How not to argue for the indeterminism of evolution: A look at two recent attempts to settle the issue. In Andreas Hüttemann, (ed.), Determinism in Physics & Biology, Paderborn, Mentis, 91-107.

Mithen Steven. 2004. After the Ice: A Global Human History, 20,000-5000 BC. Cambridge, Mass: Harvard University Press.

Morowitz Harold. 2002. The Emergence of Everything: How the World Became Complex. Oxford: Oxford University Press.

Pearl Judea 2003, “Statistics and Causal Inference: A Review”, Test (Sociedad de estadística e investigación operativa) 12(2): p. 281–345.

Penn State, 2012, « The beginning of everything: New paradigm shift for the infant universe. » ScienceDaily, 29 Nov. 2012. Web. 4 Dec. 2012.

Plummer, Paul. 2001. Vague Theories, Sophisticated Techniques, and Poor Data Environment & Planning A, 33: 761-764.

Rees Martin, 1999, Just Six Numbers: The Deep Forces That Shape the Universe, New York: Basic.

Rescher Nicholas, 1999, “How many possible worlds are there?”, Philosophy and Phenomenological Research, Vol. 59, No. 2. (Jun., 1999), p. 403-420.

Rosenberg Alex, 2005, “Lessons from biology for philosophy of the human sciences”, Philosophy of the Social Sciences 35(1): p. 3-19.

Rudra Nita, 2005, “Globalization and the Strengthening of Democracy in the Developing World”, American Journal of Political Science 49(4): p. 704-730.

Sayer Andrew, 2000, Realism & Social Science. London: Sage.

Schäfer Wolf. 2006. Big History, the Whole Story, and Nothing Less? [Review of Christian 2004 and Mithen 2004] Canadian Journal of History 41(2): 317-328.

Schlesinger, George, 1987, Nomic Necessity and Contingency. The Philosophical Quarterly 37(149): 379-391.

Schlosshauer Maximilian. 2004. Decoherence, the Measurement Problem, and Interpretations of Quantum Mechanics. Review of Modern Physics 76(4): 1267-1305.

Shalin Dmitri, 1992, Critical Theory and the Pragmatist Challenge, American Journal of Sociology. Vol. 98 Number 2, p. 237-79

Shermer Michael, 1995, “Exorcising Laplace’s demon: Chaos and antichaos, history and metahistory”, History & Theory 34(1): p. 59-83.

Silberstein and McGeever (1999), “The Search for Ontological Emergence”, The Philosophical Quarterly 49: 182-200.

Sklar, Lawrence. 1990. ‘How Free Are Initial Conditions?’, in Arthur Fine, Micky Forbes, and Linda Wessels, eds. PSA 1990, East Lansing, MI: Philosophy of Science Association, 2: 551564.

Steinmetz George, 1998, “Critical realism and historical sociology: A review article”, Comparative Studies in Society & History 40: p. 170-186.

Steinmo Sven, 2010, “History and Contingency in Political Science”, Unpublished manuscript, Univ. of Colorado Dept. of Political Science.

Sylvan David and Majeski Stephen, 1998, “A Methodology for the Study of Historical Counterfactuals”, International Studies Quarterly. 42: p. 79-108.

Tetlock Philip and Belkin Aaron (eds.) 1996. Counterfactual Thought Experiments in World Politics: Logical, Methodological, and Psychological Perspectives. Princeton: Princeton University Press.

Tucker Aviezer, 1999, “Historiographical counterfactuals and historical contingency”, Theory and Society 38: p. 264-276.

Vaidya Anand, 2011, « The Epistemology of Modality », The Stanford Encyclopedia of Philosophy (Winter 2011 Edition), Edward N. Zalta (ed.), URL = <>.

Vihvelin Kadri 2003 « Arguments for Incompatibilism », The Stanford Encyclopedia of Philosophy (Winter 2003 Edition), Edward N. Zalta (ed.), URL = <>.

Wagner-Pacifici Robin, 2000, Theorizing the Standoff: Contingency in Action. Cambridge, Cambridge University Press.

Weber Marcel. 2005. Indeterminism in neurobiology. Philosophy of Science 72(5): 663-674.

Weinberg Steven. 2001. Facing Up: Science and Its Cultural Adversaries. Cambridge, Mass.: Harvard University Press.

  1. We use the term human sciences in the sense Rosenberg 2005 does to encompass all of the studies related to society, including biology. For similar reasons we mention the “historical sciences” – biology, geology, and astronomy, where categorizing and describing are central to understanding, and more crucially, where sequences of events matter. Debate on “scientism” and the social sciences often seem to miss this point – the human sciences are scientific not because they are like physics or chemistry, but because they are like geology and astronomy, although exponentially more complex. []
  2. For example, in The Canadian Journal of Sociology: “Turner seems unaware of the fact that contingency – and therefore historical explanation – forms the core of evolutionary biology! For an authoritative statement, see Stephen Jay Gould’s [1986 and 1989].” (Bryant, 1992, p. 329n []
  3. For example, a historian writes “contingency…makes the study of history morally problematic. If the only lesson of history is the accidental coincidence of forces then what lesson do we have to draw from it? Does contingency do to the moral study of history what natural selection did 140 years ago to the seemingly moral organization of the natural world? Perhaps.” (Astor, 2010). []
  4. E.g., “Modal narratives concern themselves with contingency and determinism: with questions of possibility and necessity” (Booth et. al., 2009, p. 87). []
  5. For some of our views, see Ballinger 2008d, “Determinism and the Antiquated Deontology of the Social Sciences”. []
  6. For example, in an earlier paper (Ballinger 2008c) we list a number of examples that associate “context” with contingency, yet Bathelt and Glückler emphasize that “context is related to structure and contingency associated with agency” (2003, p. 128n). We find assertions that contingency signifies momentous changes from small effects (e.g., both Lindenfeld and Gould in Ballinger 2008c) but we find Sayer warning that “[c]ontingency is also not to be confused with importance!” (2000, p. 124). There are countless examples where contingency is associated with chance, yet Shermer stresses that “contingency…is not randomness, chanciness, or accident.” (1999, p. 218). Sayer provides a modal logic definition of contingency, yet says that it is “virtually opposed” to the popular and often academic use of the word to mean “x is contingent upon y” (Sayer, 2000, p. 28). The term frequently serves to elide clarity on all of these issues rather than commit the author to a clear position on any of them. []
  7. We use a common conception of determinism: “the thesis that a complete description of the state of the world at any time t and a complete statement of the laws of nature together entail every truth about what happens at every time later than t. Alternatively, and using the language of possible worlds: Determinism is true at a possible world w iff the following is true at that world: Any world which has the same laws of nature as w and which is exactly like w at any time t is exactly like w at all times which are future relative to t.” (Vihvelin, 2003, para. 5). []
  8. (1) is an example using the popular meaning of contingency to mean “dependent upon” and is widely used in popular and academic writing. This use is not conceptually problematic. However, it is worth noting that using “contingent upon” is a somewhat obscure way to make what is really a causal assertion. Perhaps because academic writers sometimes may prefer to avoid blunt causal assertions (Judea Pearl makes this excellent point concerning priors in Bayesian statistics, Pearl, 2003, p. 288), “contingent upon” is a useful way to obscure what would either be a more challengeable or more banal statement if put in clear causal terms. For example, if we change (1) above to read “Different kinds of redistributive policies change the effects of globalization on democratization” it is clearer, yet also now seems to be a less significant insight. The word “contingency” only served to make a somewhat obvious observation appear more important than it is. []
  9. By definition, then, even single instances of change such as Lewisian “miracles” are not permitted. Even if we contemplate them as a thought experiment, employing single small changes to somehow better comprehend a world is highly problematic. It may seem like a tiny difference for a small disc of plastic to move one millimeter one single time. But if that small disc of plastic is the button connected to a nuclear arsenal, then the change to the world is radically different than it otherwise might be. How do we know what change will and will not be truly “tiny”? Barker 2011 discusses similar problems with possible worlds semantics when possible worlds under determinism are considered. []
  10. Instead of viewing the actual world as slightly different because of miracles, one might suggest that the counterfactual possible worlds considered vary from the actual world just slightly because of tiny changes in their initial conditions and/or laws of nature. We stated above that it is doubtful those interested in the actual world are thinking of worlds with different laws/initial conditions. Changes in these may not be possible at all (Bird, 2002). Even if possible, they would most likely result in worlds so radically different than ours as to be of little interest as counterfactuals; on laws see Rees 1999, on initial conditions see Ballinger 2008b (see also the discussion of Agullo et. al., forthcoming, in Penn State, 2012, and for a philosophical approach, Sklar 1990). []
  11. E.g., in Climo and Howells 1976, Elster 1978, Hawthorn 1991, essays in Tetlock and Belkin 1996, and Ferguson 1997. Sometimes references to modal logic or possible worlds semantics are explicit, sometimes implied through reference to Saul Kripke, David Lewis, and others, and especially to Lewis (1973). []
  12. Basically we are saying here that there has been no answer to the challenge of “causal isolation” Paul Benacerraf posed in 1973, here described by Vaidya: “Realism about metaphysical modality maintains that (i) facts about possible worlds are the truth-makers for modal statements, and (ii) that possible worlds are not spatio-temporally connected to the actual world. Causal ConditionX has knowledge of P only if X bears a causal connection to the truth-maker of P. If one accepts Realism and Causal Condition, then there is a prima facie question: How can we ever know anything about metaphysical modality if we do not bear a causal connection to the truth-makers of modal statements?” (Vaidya, 2011, para. 4.2). []
  13. Thus this type of contingency could take place under “causal determinism,” now usually equated, in our view correctly, with common Laplacian concepts of determinism (e.g., Hoefer 2005, para. 2). []
  14. The similarity of this statement to the idea of a Laplacian demon is readily apparent, paralleling our point that views on contingency and views on hidden factors/determinism are in fact the same problem. []
  15. Similarly, concerning “contingencies and exogenous shocks” to a “system” Garud et. al. note “what is endogenous and what is exogenous is not given. Rather, such a demarcation depends on how actors define their boundaries at various points in time. So, what an outsider or researcher may consider to be an exogenous shock may be an activity that has been cultivated actively by the actors involved.” (2010, p. 766). []
  16. Schlesinger, in a possible worlds context as discussed in Part II, notes this connection and states “The truth-values of statements that are entirely contingent are…determined both by laws of nature and by the initial conditions prevailing in the universe.” (Schlesinger, 1987, p. 391). []
  17. There is beginning to be a unified story from primordial quantum physics through galactic, stellar, and planetary/Earth formation to biological speciation and social development. Examples of this “Big History” include Morowitz 2002, Chaisson 2001, Christian 2004, Mithen 2004, Wood 2004, and Brown 2007; see J.R. McNeill 2005 and Schäfer 2006 for overviews. []
  18. An obvious objection to this is that in the human sciences at least, this is mediated by the mind. It could be argued, however, that although they are hypercomplex mental processes are nevertheless spatiotemporal events. A short version of this is found in Ballinger 2008c and briefly mentioned in Part II above. []
  19. “Why the leopard got its spots: relating pattern development to ecology in felids”, Allen et. al., 2010. []
  20. As Jonathan Gottschall elegantly observes: “The story of the human mind is about an organ that evolved to the point of self-fascination. It deemed itself the most beautifully complex phenomenon in the known universe, turned inward, and decided that its greatest challenge was to plumb the mysteries of its own nature. These mysteries cannot be scientifically addressed without the storytelling discipline of evolutionary psychology. Without EP we can perhaps develop a solid description of what humans are. But we can’t get to “why?” To stop telling and testing just-so stories would be to forfeit hope of answering the oldest, deepest, and most compelling questions about our species: Who are we and how did we get this way?” (Gottschall 2012). []
  21. Confusingly, these are frequently conflated in a single work: “Path dependence is the application to economic systems of an intellectual movement that has lately come into fashion in several academic disciplines. In physics and mathematics, the related idea is called chaos — sensitive dependence on initial conditions…In biology, the related idea is called contingency” (Liebowitz and Margolis, 1995, p. 33). Even when not explicitly conflated as in the example above, chaos, path dependency, chance, free will and the other confusing concepts we find associated with contingency are so often found in such poorly defined association with one another as to encourage their conflation, e.g.: “the pragmatist sensitivity to indeterminacy, contingency, and chaos.” (Shalin, 1992, p. 238) and “the event structure displays the crucial role played by human agency, contingency, and path dependency” (Griffin, 1993, p. 1117). []

Vous aimerez aussi...

1 réponse

  1. A version of this article with footnotes instead of endnotes (PDF) is available here: Contingency is just so.

Laisser un commentaire

Votre adresse de messagerie ne sera pas publiée. Les champs obligatoires sont indiqués avec *