The following is an excerpt from a draft commentary I’m working on, going through Ellul’s Propaganda chapter by chapter, section by section. Have I talked about bouncing off Ellul’s work before? Yes, I have. Why am I returning to it now? I wanted to refine my thoughts on persuasion and propaganda, so I thought the most expedient way was to bash my head against a work I know I disagree with. Along the way, my critique will (and already has) push me to expand my reading to include sources I may not have originally considered.
I am reworking the commentary now; it hadn’t progressed too far by this point, but I decided I needed to rethink some things. But I have decided to post my original response to his section Total Propaganda under Chapter 1.1: External Characteristics. This section might survive in the final commentary or it might be significantly re-written, but a major impetus for me starting on this commentary was to oppose Ellul’s concept of total propaganda with my concept of the propaganda machine.
Ellul begins this section with this quote: “Propaganda must be total. The propagandist must utilize all of the technical means at his disposal—the press, radio, TV, movies, posters, meetings, door-to-door canvassing. Modern propaganda must utilize all of these media. There is no propaganda as long as one makes use, in sporadic fashion and at random, [of them]…. Each usable medium has its own particular way of penetration—specific, but at the same time localized and limited; by itself it cannot attack the individual, break down his resistance, make his decisions for him.” While this is a continuous quotation, it will do us best to deal with it in two parts, with the second part beginning at “Each usable medium”. The reason for this is while the entire statement is false, the two parts are differently false. In the first part, the failure is that he has a binary view of propaganda; in the second, it is that he attaches specific semi-mystical qualities to media which it does not warrant.
In the first place for this first argument, propaganda does not need to be total. The chief and most important reason for this is to permit the study of propaganda. What Ellul suggests by this statement is that we can only study successful propaganda as propaganda, since by definition any unsuccessful propaganda was not total, did not achieve its ends. By defining propaganda as “propaganda which has succeeded”, we’ve created a circular definition. In order to avoid this we have to be able to say “these are elements of propaganda, which are in themselves propaganda-like”. Rather than restricting our definition and thereby creating the specter of propaganda which cannot be defeated, we expand the definition to include “partial propaganda”, as said before. Propaganda is just communication from a particular point of view, so it is generally encountered in a partial, non-encircling way.
In the second place, we can recontextualize Ellul’s “total propaganda” as “the functioning of the ideal propaganda machine”. That is to say, when Ellul describes total propaganda, what we should take this to be is a description of the output of an ideal propaganda machine. It isn’t the case that all propaganda must be like this, but a total and encircling propaganda is a kind of apotheosis of the effect of propaganda.
To understand what a propaganda machine is, we first have to understand the basic structure of communication. At either end of a communication we have an individual: in a two-sided communication, one will be the speaker and the other the listener. Each individual has a variety of traits. To take from an essay by Harold Lasswell (one of Ellul’s sources), “The structure of any personality may be conveniently classified according to ‘identifications,’ ‘demands,’ and ‘acceptances.’” These are all traits which have varying levels of importance to that personality/individual. For importance, we can give a generic definition such as “the degree to which the condition of a thing is considered worth maintaining”. If an individual’s important trait is threatened, they are likely to take some action to safeguard or repair that trait. In fact, we could say that this action is what demonstrates and calcifies the actor’s possession of that trait both in view of the actor and in the view of observers.
If a speaker sends a message (i.e. communicates) to a listener, that message will bear on an idea. Traits and ideas are the same thing, being intangible conceptions, except that traits are those considered to belong to a person while ideas are considered to be free. Regardless, both have a level of importance to an actor. The interplay of the message’s content and the importance of the central idea are what result in the reaction of the listener.
A straightforward example of this is death-coercion, or coercion through the threat of death. The central idea here is the life of the victim. When a message is delivered to the victim which highly endangers their continued living, they are highly likely to take an action based on the content of that message rather than ignoring the message.
Of course, it is true that communications may refer to multiple ideas, and that people don’t always consider just one thing when making decisions; the model can be made more complex, but this is the most basic version. The point of this basic model is to establish that communication is not simply active on one side, it requires an active comprehension.
Lasswell states “An act of communication between two persons is complete when they understand the same sign in the same way.” This is incorrect. I believe it is based on the idea of a single truth which is being expressed, the idea of a correct meaning. But there are no correct meanings. An act of communication is complete when it is comprehended by the listener, regardless of how it is comprehended. Verifications might be communicated back and forth to attempt to establish that each side does have the same understanding, and it is possible for two sides to have the same understanding; nevertheless, the Two Generals’ Problem is unsolvable. Therefore it cannot be the case that two people must know they have understood the sign in the same way, otherwise it could never be proved that communication has taken place.
When we remove the idea that there is a specific way to read any sign, we also remove the idea of any direct link between the intention of the speaker and the reaction of the listener. It’s impossible to say from a single piece of propaganda what the broader goals of that propaganda is. A highly anti-Black group might put out posters etc. full of anti-Black sentiment, but even a statement like “Kill the Blacks” does not give the entire campaign. To some it will say that Black people are dangerous; some of these will say they must defend themselves, others will say they must flee. To some it will say that Black people are degrading the society; again, there are several conclusions which could be drawn from here. Such a statement is clearly propagandistic, but it doesn’t constitute a developed pressure in a particular direction.
A propaganda machine is an organization for coordinating propaganda towards a particular end. Ellul’s conception of total propaganda does require a propaganda machine – “The propagandist must utilize all of the technical means at his disposal—the press, radio, TV, movies, posters, meetings, door-to-door canvassing.” – but he does not specify a specific machine or coordination here (I believe he does later, and we’ll deal with it then). In part, this is because of his mysticism of the mediums, each medium having its own preferred targets and operations, etc. He doesn’t place importance on the speaker of propaganda but on the methods used, and therefore he doesn’t seem to theorize that propaganda can be other than state propaganda. He will later go into ideas of political and sociological propaganda which only support my argument here: Ellul’s propaganda is always delimited, always top-down, from the very top of the society.
I will summarize Ellul’s conjecture about mediums by saying that he believes certain types of people engage in certain types of media and thus they are only reachable through those media. I think this highlights exactly what this “medium mysticism” is to me. Because in every example that I have come across so far, Ellul refuses to detail how a piece of propaganda acts upon a person. He says “[a] word spoken on the radio is not the same, does not produce the same effect, does not have the same impact as the identical word spoken in private conversation…” Why? It will seem obvious that radio speech and private speech act differently, but it is not clear that this is true, especially when we talk about propaganda. How can we say a priori that radio speech and private speech will “naturally” reach different publics or “naturally” have different effects on them?
Ellul appears to believe that propaganda simply does affect people on contact; there is no other operation, there’s no appraisal, one simply experiences propaganda and is affected by it. The “job” of the propagandist is less to convince and persuade than to simply reach. Mediums differ primarily by the type of public they reach and the content they can carry. Nothing else about the process is important. But this is misguided.
It’s striking to me, but not unexpected, that Ellul offers no theory or model of content in his work. It does not appear in the list of subjects he explicitly excludes, either. Whereas Lasswell does develop a rough semiotic which allows him to at least question how each side understands a communication, Ellul does not. He takes it for granted that propaganda is understood and in essentially the same way by the speaker and the listener. Again, this is in part a consequence of his need to consider propaganda as total from the beginning.
But medium mysticism must be replaced, and it needs replacing because it does perform a function here: it tells us why a listener would respond to a piece of propaganda. Ellul makes note that psychological propaganda works very infrequently, in small percentages; any research on election campaigning will tell you something similar. Medium mysticism allows Ellul to say something like “while most people are not susceptible to most propaganda, most people are susceptible to some kinds, and therefore total propaganda can proceed by utilizing the complete set of mediums”. The problem with Ellul’s medium mysticism is that it’s incorrect, for the reasons I’ve just laid out.
Instead of succumbing to this mysticism, I will replace it with a different idea: the listener ascribes a certain “relevant ascendancy” to the speaker, and this influences how the listener understands the message. For “relevant ascendancy” it is acceptable to substitute the word “authority”, but this has to be understood as “relevant ascendancy” rather than “power”, “creation”, etc. It refers not to legalistic authority but social-cultural reliability; it maps onto expertise more than command, to prestige more than legal privilege. A speaker does not have the same relevant ascendancy for every message and idea, and each listener will perceive the same speaker as having a different level of relevant ascendancy.
In this case, we can say that a listener is persuaded by a speaker’s message if the listener is convinced by the content of that message, as well as the authority they confer onto the speaker, that the message has some significant bearing upon a central idea which the listener views with some level of importance. If this has happened, then propaganda has been successful.
The reason for the multiplicity of messages and mediums being used by the ideal propaganda machine is not that each person requires a specific medium, but that each person has holds different ideas to be important and has different perceptions about possible speakers. These perceptions are not inherent to the mediums; media works to constrain and adapt messages, but they don’t alter such messages with a propagandistic purpose. If one has a video message that they wish to play on TV, but then take the audio from that same message and also play it on radio, this hasn’t created two dramatically different pieces of propaganda which will naturally affect different people. Instead, it’s the same type of person who would be affected by that similar message; what differs here is their access to the media, not the type of medium, and even the fact of access is a bit overblown as it regards propaganda. That is to say, it is tempting to conclude that because televisions have a certain cost, that TV buyers therefore constitute a specific public with its own whims; this isn’t true. Differences between social groups do certainly exist but it is far too cute to try to boil it down to “newspaper readers think this, while radio listeners think that”.
In addition, the propaganda machine does not need to encircle anybody. As I’ve just laid out, propaganda works when it’s successfully convinced a person. It does not require being part of the mass or being surrounded on all sides. The purpose of the propaganda machine is not to produce enough propaganda to overwhelm a person or even enough to saturate the society: it is to create particular ends. War propaganda may rely on the underlying culture of a society, but its goal is not to create a totally-enclosed propagandized subject, its goal is to convince people to go to war, and these are completely different propositions.
Of course, it is true that people in the past have attempted to create totalitarian states through propaganda. This can’t be denied. But it is too much of a leap to go from there to say that all propaganda, or even all propaganda machines, are totalitarian in nature. The significance of a propaganda machine is not that it has a total coverage of any kind; this is in fact very rare, relative to the widespread nature of propaganda in general. The significance is that it uses many different mediums & methods and that these are coordinated. It isn’t through an analysis of any particular piece of propaganda that we can infer an overall intent, but we can infer the intent of a propaganda machine by analyzing its output as a whole. Since we are considering propaganda to be a view of communication rather than a new kind of object with a strange power, a propaganda machine does not need to cut off its targets entirely, it only needs to speak to them in ways that manage to persuade.
In order to make the distinction between propaganda as a viewpoint on communication and propaganda as the output of a propaganda machine, I will refer to the first simply as propaganda (or broad propaganda) while I will refer to the second as propagandization.
This is not to say that there is not a kind of encircling effect happening in the modern world, in which propagandization plays its part. But as I expect to elaborate more on when we speak on Ellul’s notion of sociological propaganda, it is here that we have to differentiate propaganda from non-propaganda. If propaganda is communication viewed in a specific lens, then non-propaganda is non-communication. In the section introducing this idea, Ellul says this: “Through the medium of economic and political structures a certain ideology is established…”. And later: “Sociological propaganda springs up spontaneously; it is not the result of deliberate propaganda action.” He locates this “in advertising, in the movies (commercial and non-political films), in technology in general, in education, in the Reader’s Digest; and in social service, case work, and settlement houses.”
What I see in this is a motion that Ellul doesn’t call attention to: the origination of sociological propaganda. He calls this propaganda spontaneous, as if it comes from nowhere. But clearly if we take sociological propaganda to be the propaganda actions (movies etc.), then there are the structures which lay behind it that are important; and if we take the structures to be social propaganda, then there are the social and economic conditions in the “pure” state which underly the structures. It is not easy to separate the structures from the conditions which birth them, but both can be examples of non-propaganda: conditions are never propaganda, while many structures are not in-and-of-themselves propaganda, or rather, they more usefully viewed as not-propaganda.
If propaganda is communication intended to persuade, we have to assume there exists something akin to “real conditions” which individuals are being persuaded about. If I want to persuade someone they should buy lemonade, I might need to bring up the fact that it’s hot outside and a lemonade will cool them down, or that they like sweet things and lemonade is sweet. The sweetness of lemonade and heat of the day are not propaganda. The cultural idea in the United States that lemonade should be sold on hot days is certainly propaganda (though not necessarily concerted propagandization). But is the existence of the standard of buying-and-selling a kind of propaganda? That is an question which isn’t easily answered in an agreeable way, but it certainly seems to be of a different order than the connection between lemonade and summer.
I would therefore say that as far as knowable things, notions, or ideas in relation to the concept of propaganda, we have to split them into at least the following three categories: conditions (which are not propaganda), broad propaganda, and propagandization.
Ellul’s distinctions between direct and indirect propaganda, covert and overt propaganda, seem immaterial to me. Direct propaganda maps onto propagandization, and indirect propaganda to broad propaganda; there’ll be a chance to delve more into this later. The distinction between covert and overt is a conventional moral categorization, not really an analytic one. It approaches usefulness if you simply link covert with indirect and overt with direct, but in that case it’s just a restatement of terms we already have. It does allow Ellul to go on a moralizing rant which is his prerogative.
I gather that these sections might be popular with some of Ellul’s readers but they simply hold nothing for me. I think it’s because what I’m looking for in this work is analysis, and these sections are primarily polemic. Is there an analytic value to pointing out that some kinds of propaganda work together to create a combined effect? Not in the way Ellul does, in my mind; as if it is a shocking result instead of an obvious one. The overt gives cover to the covert in all things; it’s an elementary lesson in warfare, for instance. It’s perhaps the case that people were much more naive about propaganda in Ellul’s day but, given the amount of writing Ellul himself was able to draw on, I doubt that conclusion.
I find it interesting that Ellul begins his last push in this section by saying “[l]et us give one last example of this combination of differing types of propaganda,” as if the specific distinctions are not themselves important. He begins the section’s last paragraph with “Although this distinction is not altogether useless…”, whereas he’d earlier said “[w]e must also distinguish between covert propaganda and overt propaganda.” It is as though, along the way, he lost his conviction in these distinctions. At first the distinctions must be understood, apparently on their own terms; then the distinctions are to be seen just as examples of the idea of distinguishing different types of propaganda; and finally the distinctions are essentially immaterial. But it’s just as well, given that the distinctions are not well-founded in the first place. I will not go into a detailed dissection of each because I have already made my broad argument as to why these distinctions are immaterial, and Ellul himself concludes that no modern propagandist would use only one or the other kind, regardless of what distinction was being made. We will leave aside his implication that propagandists of the past would have been able to use only one form or the other.