Authoritarianism is a specific thing, briefly the idea that submission to the will of another is an intrinsic good and any independence and autonomy in conflict with the commands of an authority an intrinsic evil. But opposition to authoritarianism is not a specific thing. Any compromise of the absolute, unconditional nature of authority constitutes anti-authoritarianism, from the absolute opposition to anyone ever conforming herself to the will of another under any circumstances to merely the idea that a subject has at least the right to protest an order, even if she is ultimately forced to comply over her objections.
Anti-authoritarianism has been around at least since the beginning of the 18th century, and arguably goes back to ancient Greece. In the West, anti-authoritarianism has, in one form or another, been the predominant political philosophy since the late 18th and early 19th centuries bourgeois revolutions, and goes by the general label of "liberalism". Just because an general component of political philosophy is predominant doesn't mean it's unchallenged. That the 20th and early 21st century "conservative" revolutions in the West are indeed seen as radical revolutions seeking to overturn the liberal status quo is by itself strong evidence for the present predominance of liberalism.
Denying the underlying narrative of authoritarianism — that authority must be obeyed just because it is an authority — demands a counter-narrative. should one person ever subordinate her immediate will — even a little — to the will of one or many others? And if so, when, under what circumstances, and to what degree?
At least under present circumstances, we have to say that yes indeed, sometimes an individual must subordinate his will to the will of others. This view is not authoritarian, because it denies that a person must always, unconditionally submit: sometimes is not always. We are forced by physical necessity to cooperate; there are benefits we must have to survive that can be gained only by working for a mutual goal. Furthermore, we know there are situations such that:
- If everyone in a group cooperates, everyone is better off than if no one cooperates
- If everyone else cooperates, any individual is always better off if he does not individually cooperate
One good way of managing this situation is creating the ideal of voluntary cooperation. Voluntary cooperation consists of the following premises:
- An individual can join any group with the consent of its existing members
- By joining a group, an individual agrees to cooperate for the mutual benefit of the members of the group. By implication, a person will join a group if and only if she believes that cooperation will be more beneficial than independence or membership in any other group
- If an individual does not believe that cooperating within the group is in her best interest, she may leave at any time, taking her share of the physical assets of the group with her.
- If the rest of the members of a group do not believe that a specific member is cooperating, they may expel that member from the group. (Technically, the rest of the members can leave the original group and form a new group without the offending members.)
However, actual voluntary cooperation is not presently achievable.* The rub is (3). Obviously, if an individual were forcibly constrained to be in a group, we could not call it voluntary cooperation; an individual in a group could be exploited rather being guaranteed to enjoy the mutual benefits of group cooperation. But also, if it's not possible for an individual to survive and prosper individually, (3) is, in principle, equally compromised. It's even more compromised if there are a limited number of groups, and especially if there are significant costs of forming and/or joining a group.
*I'm not saying (at least not yet) that any group actually believes or proposes that actual voluntary cooperation is or is not achievable.
It's not at all clear that actual voluntary cooperation would ever be achievable. Regardless of group participation, it's still physically possible for individuals in a group to physically coerce individuals in other groups. It's not at all clear that equality of coercive force between groups would emerge from the principles of voluntary cooperation. Indeed it seems clear using perturbation theory that inequality of force would emerge. Suppose one group had a slight advantage of coercive force over another. The first group could then recruit new members by offering the mutual benefit of coercively exploiting the second. To the extent that the size of a group offered it a coercive advantage, a small disequilibrium would have a positive feedback effect magnifying rather than dampening the disequilibrium.
Actual voluntary cooperation might also not be economically possible under any circumstances. Presently, an individual cannot even survive completely on his own; membership in some group is absolutely compelled by physical necessity. But even if individual survival were possible, it is not at all certain that living a "good" life would ever be possible as an individual. It might well be the case that people would always define a "good" life as enjoying benefits available only by cooperation. Even if every person actually had a Mr. Fusion and a replicator, it might be that interplanetary or interstellar travel was considered necessary to live a full and complete life. In much the same sense, even today we define a good and dignified life as not just having enough to eat and sufficient shelter to survive a snowstorm, but having a clean and comfortable dwelling, healthy and nutritious food, access to not just life-saving but also life-enhancing medical and dental care, access to communications and media, and other benefits over and above sheer survival.
But it's not at all necessary that voluntary cooperation be actually achievable to be a useful idea. Even though for example Newtonian mechanics is an enormously useful idea for modeling the behavior of physical objects on the surface of the Earth, even though in practice air resistance and gravity make ideal behavior actually impossible. In much the same sense, we can use voluntary cooperation as a theoretical ideal to measure the quality of an actual social system: We can define a system to be good to the extent that it creates an outcome similar to the outcome we would predict from actual voluntary cooperation. The abstract ideal of voluntary cooperation gives us an objective measurement of the quality of a social system. (Note that having an objective measurement does not compromise meta-ethical subjective relativism: we are still making an arbitrary decision to privilege voluntary cooperation as our standard of measure.)
All "liberal" political philosophies explicitly or implicitly reference the abstract ideal of voluntary cooperation as the standard of measure of the quality of an actual social structure. Even crypto-authoritarians must at least be crypto- and give lip service to voluntary cooperation as a standard of measure. All arguments over political philosophy consist of either what benefits would result from the abstract ideal of voluntary cooperation or to what degree the benefits of some existing or proposed social structure would actually match the theoretical benefits of voluntary cooperation. Even its proponents of capitalism assert that bourgeois democracy best matches the benefits of voluntary cooperation under present physical circumstances. But note that any social system that is not actually voluntary cooperation is not actually voluntary cooperation; at best it is something other than actual voluntary cooperation that more or less resembles voluntary cooperation. We can at best say that a social system implements effective rather than actual voluntary cooperation.
Anarchists are not at all explicit about what objectively constitutes "institutional" or "hierarchical" authority. Merely saying that the authority of expertise does not constitute hierarchical authority is insufficient. It's also somewhat misleading: one can use expertise to exercise actual power over others. Physicians, for example, are presently using (or allowing others, — insurance companies — to use) their expertise for truly staggering economic exploitation.
There are two ways, then, to infer a meaning of "hierarchical" or "institutional" authority to which anarchists object. The first is any coercion or necessity that causes the outcome of a system to deviate substantially from the abstract ideal of voluntary cooperation. There's nothing "wrong" with this definition; the only problem is that it doesn't at all differentiate anarchism from liberalism in general.
The alternative reading is that hierarchical or institutional authority is any coercion that deviates from actual voluntary cooperation. This reading is more problematic.
One way to achieve effective voluntary cooperation is to privilege some individuals with the power to make sure that other individuals do not actually behave contrary to the principles of actual voluntary cooperation. For example, we privilege police, courts and prisons to ensure (at least in theory) that rapists and murderers do not "infiltrate" the group of non-raping, non-murdering people. We do so by technically violating the principles of actual voluntary cooperation by forcing rapists and murderers to join the prison population group. While we are (in theory) violating the principle of actual voluntary cooperation, the results of this violation implements effective voluntary cooperation: the outcome of a system where we violently coerce rapists and murderers resembles a system where the group of non-raping non-murdering people could effectively exclude a rapist or murderer.
Anarchists make (as best I can discern) a not-at-all stupid argument that privileging people in this way cannot result in effective voluntary cooperation. Privileging some individuals to maintain effective voluntary cooperation will result instead in that group coercing others for their own exploitative benefit. I'm not at all confident that this argument is false or unsound. But what's the alternative?
The only alternative I can see is that the social ideas prevalent in the population are such that effective voluntary cooperation occurs without any individuals being privileged, even putatively to coercively maintain effective voluntary cooperation. The problem comes in how to make these social ideas prevalent in the population, and how to keep them prevalent.
Exhortation and positive propaganda, i.e. trying to convince a lot of people of the value of these social ideas, does not seem sufficient. Again, we can use perturbation theory to criticize this method. Suppose almost everyone really did have all the social ideas necessary to effect voluntary cooperation. But suppose that a few people did not have these ideas. Because our society ex hypothesi effects voluntary cooperation, these few people can freely — without violent coercion or coercion by necessity — leave the group of people with the "right" sort of social ideas and create a group with the "wrong" sort of social ideas. Suppose these "wrong" ideas included the ideas of within-group cooperation, but rejected the idea of inter-group noninterference: they work together as a group, but think that it's OK to coerce other groups. Voluntary cooperation does entail any "natural" or systematic way this group would not grow arbitrarily large, large enough to actually benefit from inter-group coercion, a relative benefit not availed by other groups.
Note that this argument is not the naturalistic fallacy, which is the fallacious conclusion that because some state of affairs actually exists, it is therefore preferable. This is an argument from "human nature", but it's an argument from a deep character of human nature: our ability to form novel social ideas. It is not an argument that the present state of human social ideas constitutes an ineluctable, physically necessary condition.
It's also not a slam-dunk argument proving that the second reading of anarchism is definitely impossible. We simply do not know what's possible and impossible; we only know what is or is not probable or plausible given definite initial circumstances. It may be the case that some set of social ideas, and perhaps some method of how human beings transmit and select against social ideas, would effect voluntary cooperation without either privileged coercion and by somehow "naturally" dampening dissent from shared social ideas.
But "Anything's possible," is not by itself a very persuasive argument. If the anarchists (of the second reading) ever figure out how — even in theory — to implement a society that can effect voluntary cooperation without somehow privileging someone to maintain cooperative behavior, I'm all ears. But until then, I'm going to focus my energy on not eliminating but improving social institutions and authority structures to better serve the needs and mutual benefit of all humanity.