Thoughts After Reading an Émile Durkheim Excerpt for a Course


One of the things I enjoy about Durkheim so far is the way he uses notions of both space and time to analyze society. It is rather ingenious. It does seem like the density and number of nodes on a social network have an effect on the way time is socially experienced across the network, as well as other features of society and how individuals experience society. Put in other terms, Durkheim links network topology (looking at a network as a geometrical object that can experience displacement of its components while preserving certain properties) with social ontology (the social being of the individuals that form the nodes of the given social network). I’d like to expand on this insight, since so far at least Durkheim is not incredibly clear about his reasoning process from quantitative population measures and network size to qualitative differences in solidarity. What he does make clear is the relationship between qualitative differences in solidarity and the sophistication of the division of labor. It is this which makes me not fully buy into Durkheim, and, I think, if I go over why this doesn’t make me buy into it in detail, it may reveal this comes from some of my sympathies for Marx.

So maybe I should think abstractly about social networks for a moment. For example, if you have millions of people you socialize with, each of whom know at least a subset of those who you socialize with, even if its to varying levels, one can say that your social network is “high density.” Perhaps we could measure density in terms of how often network links intersect with each other, and thereby how many times an intersection must happen (wherein the intersection is not yet another individual, and is thus not a node insofar as nodes are specified to be individuals in our analysis here). Or, perhaps, in terms of how many links are likely to radiate from any given node in the network. In any case, a high density network is going to affect each node’s experience of time insofar as each node’s traits remain constant (or, more generally, to the extent that nodes are constrained in their use of resources and their rate of capacity to reproduce themselves before they’re completely spent out of existence). That is to say, a high-density social network would make agents in them experience a higher pace life since it may require them to divide their time, however disproportionately they may do so, and it would also make agents in them experience higher risk in decision-making. Both of these are because:

  1. Each node has to divide their time in a way correlated with the amount of links radiating from them, leading to a loss of integrated social information by which to make decisions, increasing uncertainty about the general reception of their personal/social decisions;
  2. Each given node is more saturated with information that is of personal relevance (whether negative or positive), which means that cultural symbols lose consistency of meaning across groups and the likelihood of miscommunication across social groups is thereby high. Society can here then be said, not only to “speed up” since social formations become more ephemeral/flexible with higher density, but higher density can also put pressure on nodes in the network to the degree that there may be an incentive to meet social status goals more quickly or as quickly as other individuals.

Yet, Durkheim’s interpretation of this breaks down right here. It’s not enough for society to be high-density, in order that it speed up or accelerate. This already presupposes a generalized social status factor that requires a high degree of competition, due to its object being scarce (whether naturally or artificially). This is not a given. An alternate possibility is for high-density networks to incentivize the growth of each individual’s capacity for manipulating abstraction as well as each individual’s capacity for metacognition (or, capacity for entertaining a multiplicity of perspectives/interpretations of behavior).

In the same respect, it could then disincentivize them from relying on fight-or-flight, fear, etc., as sources of problem-solving. This is because the necessity of these skills seem to be latent in the fact that those individuals have to deal with more complex social networking (although, again, this is mediated by status-seeking). Further, Durkheim seems to see this “civilizing” force as, in retrospect, promoting individuation (i.e., increasing the average uniqueness of traits across individuals). But he never asks whether this individuation is on the individual’s own terms; he seems to implicitly think so since his language for describing what happens in this individuation is that society leaves a space open for the development of individual powers. He doesn’t say that society produces that uniqueness. Rather, for Durkheim it seems to be a side-effect of the space left open, to members of society, for independent action. This way of characterizing it may be a bit naive, for if indeed society also tends to speed up or accelerate with this process, it’s not clear where Durkheim’s assumption of agency comes in. Is it not the case that agency may be decreased given a loss of huge amounts of integrated social information available to the individual?

Even Durkheim’s point about specialization and the division of labor may not be straight-forward if its justified only on the basis of properties of the network. To illustrate: The other property of the network we can look at in terms of space is the number of nodes in the network. It seems obvious that, since any given level of density in a social network is also constrained by the quantity of nodes/agents in the network, a high number of nodes/agents in society is necessary but not sufficient for high network density. What, then, is the more direct impact of a high population of societal members on that same society, if it’s not necessarily high network density? Well, with higher population, a higher range of differentiation is possible simply because there is a higher number of possible permutations that may act as variables related to particular traits that may be manifest in individuals, and because there is a higher variability of possible distribution of individuals in that society across a whole geography. Higher population, then, is likely to increase diversity.

However, here’s where Durkheim’s optimism falls apart a bit: Durkheim interprets this diversity to mean specialization. The previous point I mentioned shows this to be a non-sequitor, as the skills that a high-density network can promote may allow for incredible ease at multi-tasking, integrated application of different skills as a single person, and an over-all improvement of performance across various domains. While the level of skill or performance in some things would still be higher than others, other skills and performance can nonetheless be high enough that tasks requiring them do not have to be outsourced to other labor so that the given project undertaken is successful. The relevant variable being what “successful” is, but it will involve an average understanding of success by that society that may not need to absolutely emphasize productivity (output per unit time, in money or product) or efficiency (trimming of labor extraneous to the basic expected average use of the product).

Even if we grant that the domain-general skills a high-density socially networked society demands of individuals need not prevent specialization even if may make it less necessary in more general terms, there’s another issue with Durkheim’s assumption that diversity means specialization: he lumps, or conflates, all division of labor as essentially the same. This makes his parallel between the development of mechanical solidarity into organic solidarity, and the development of low specialization to high specialization, quite suspect, even if we take it to be a logical/ontological claim (a claim about one thing’s logical dependence on another, or one thing’s dependence on another for its existence) rather than a historical claim (a claim about what came first in a temporal order, in actual fact). How is it suspect?

Well, given what has been said so far, it seems reasonable to say that mechanical solidarity can hold, internally, a high degree of specialization if it is extremely strict, as all mechanical solidarity requires is that individuals in the society be highly integrated into performing a single function at the societal scale. We see this in some ants. What it can’t incorporate is a flexible and highly dynamic allocation of individuals into particular divisions of labor. Organic solidarity, on the other hand, could probably internally hold both sorts of division of labor. In fact, we see this in how some societies are constituted by some rather essential social conflict due to the disproportionate significance given to certain divisions of labor, or in how the production of a division of labor may be largely related to culturally specified roles more than they are necessary to the material provision or production of goods in more general terms (this role having material necessity only to the extent such a contingency itself affects the politics of distribution of wealth and of the distribution of the returns on production). On the other hand, if we see history in terms of forces of production, as Marx does, we can see how this “cultural contingency” is not entirely a contingency, but a matter of the relative development of the forces of production from one point in history to another given both scarcity and the inextricability of consumption from the development of unified culture out of inconsistent norms and values.