This is a preliminary attempt to unpack Sicart's concerns and rephrase some specific concerns in a way that makes sense to me. It's difficult to evaluate the argument as presented. What I understand as proceduralism and procedural rhetoric is considerably broader than what Sicart appears to, and the vast majority of writers I'd associate with a proceduralist approach don't seem to be mentioned at all, nor their texts engaged with. One can't do everything, but a strong interpretation of this paper would have it aiming to refute not only Bogost but also several decades of other work on expressive and rhetorical use of procedural artifacts, which seems implausible for it to do without discussing them. Thus it must have some more limited scope in mind; and I think as a result, the binary opposition it sets up between proceduralism and play-centrism is too simple.
From the title and some of what's hinted at in the body, the article at first sounds like a general attack on a broader kind of proceduralism: an approach to computer-mediated interactive experiences that considers the key affordance of the computer to be its capacity for executing procedures. Designers coming from this perspective often consider the key method of enabling new experiences to therefore be a focus on what one can do with procedures (either new ones, or old ones deployed in new ways).
However, despite a mention of Janet Murray on that point, that doesn't seem to be Sicart's key target, or at least he doesn't spend any time in the article targeting that broader kind of proceduralism; hence one finds no engagement with Edward Ihnatowicz, Simon Penny, Chris Crawford, Michael Mateas, Noah Wardrip-Fruin, or any other people I'd put at the core of proceduralism, nor of papers arguing explicitly for a focus on procedures as a game-design agenda.
We might limit to a more specific concern about proceduralism: opposition to procedural rhetoric, the idea that one can deploy procedures for rhetorical purposes. But again this is hard to see as a full-on attack on the idea as I see it; it must be meant in some more limited way.
The core idea of procedural rhetoric is that arguments can be encoded in processes or simulations. Its precursor is the proceduralization of theories into simulations: by the 1940s, some scientists were arguing that, for example, a psychological theory of how short-term memory works would be better implemented in a simulation rather than described in prose or diagrams—a simulation being a more precise and testable theory, with fewer places to sweep murky details under the rug.
But scientific modeling is only one specific type of procedural argument: a set of procedures coupled with the claim, "I believe these procedures accurately model how [a thing] works". Other rhetorical modes are possible. The most obvious is a caricatured model that exaggerates how something works rather than accurately modeling it, serving as what you might call a procedural satire of a subject. The earliest prominent procedural satire is probably Robert P. Abelson's "Goldwater machine", which simulated 1964 U.S. Presidential candidate Barry Goldwater's view on current events in a piece of artificial intelligence software that would always find a way to blame communists for absolutely anything. Procedural satires proliferate on the internet as well, as proceduralized critiques of academic writing styles, revolutionary manifestos, and much more.
How exactly these satires function is an interesting and probably under-theorized point (though see an article by Mateas on how "computers are fundamentally meaning machines"). One view is that they embody critiques that are already themselves procedural. Summarizing Barry Goldwater's ideology as "always blame the communists" is a claim about how Barry Goldwater operates, in particular that he operates according to a simple script (to use Abelson/Schank terminology) roughly as follows: "Step 1: consider the issue; Step 2: blame communists". We could just claim that in prose. We could give a few examples of how Barry Goldwater might blame communists for various things in sort of comical ways. We could produce a flow chart—a common technique in editorial cartoons—that summarizes Goldwater's thinking process (all arrows lead to "blame Communists"). Or, taking Abelson's approach, we could produce a satirical simulation of Goldwater's thinking, where we can feed it new examples and it'll return new ways of blaming communists for the ills therein.
Thus from a background of artificial intelligence or electronic literature, the consternation caused in some circles by Bogost's 2007 book Persuasive Games, which proposed that one could use simulation games' rules as rhetorical devices, was surprising. If anything, I was expecting the opposite criticism: "sure, that's right, but this isn't new". That would've admittedly been a bit unfair, but certainly the basic idea that one can deploy procedures encoding a viewpoint, for rhetorical purposes, didn't even really raise eyebrows in my circles.
I don't see this as a game-specific problem, however, nor as a critique of the concept of procedural rhetoric itself. Instead, I read it as opposition, aesthetically and/or politically, to certain kinds of unsubtle, didactic rhetoric in general—of which unsubtle, didactic procedural rhetoric is one variety among many.
Serious games and training games often aim to replace or augment textbooks and videos, many of which make quite unsubtle, didactic use of more traditional kinds of rhetoric. Overtly propagandistic procedural rhetoric, like that employed by some games from Molleindustria, is unsubtle in the way propaganda is always unsubtle, in games no less than in pamphlets or films (though Molleindustria is certainly quite self-aware about it).
If anything, the fact that one can produce a piece of propaganda via exaggerated, biased, and/or caricatured game or simulation rules, rather than solely through exaggeraged, biased, or caricatured textual or audiovisual elements as in a propaganda leaflet or film, seems to be a vindication of the concept of procedural rhetoric. If indeed procedures can function as rhetorical devices, then it should be no surprise that the devices can be employed in all the same heavy-handed ways as traditional rhetoric, producing propaganda, moralistic fables, or unsubtle satire. This doesn't necessarily mean that they must be deployed in such ways, however. Nor do I think that the culprit in most cases is game-specific: didactic educational games mainly reproduce the educational ideologies of the institutions that fund and deploy them.
Put differently, at times Sicart sounds like he's arguing that "proceduralism" has made a philosophical error, in that it erroneously claims game rules encode more meaning than they really do. But his concerns about instrumentalizing the player to convey didactic, moralizing messages seems to rest on the opposite view: that some games really do foreclose any real role for the player in meaning-creation. Sicart would prefer other games to be made in which that wasn't the case. But I think that's better phrased as a disagreement with certain uses of rhetorics, whether procedural or otherwise, rather than as a critique of the concept of procedural rhetoric.
To turn to the (similar) art-games question by way of non-game art: There is a sculptural work in the Copenhagen harbor in which several ghostly white figures are marooned on a platform offshore. The artist, an immigrant to Denmark, designed it as a statement about the difficulties immigrants encounter when integrating into Danish society. The marooned sculptures symbolizing integration difficulties is arguably not the world's most subtle metaphor. The same could be said of a number of art games in which the movement and interaction of different components is said to represent a concept or entity onto which it maps fairly directly. But again I don't see this as a particularly unique problem with proceduralist art-games, just with a certain kind of direct-symbolism in art works. (There is also a large related discussion, hinted at by Sicart, about the role of artists' statements that explain what a work is "supposed to mean".)
I'll submit this as a possible position:
"Meaningful games" should not be modeled on rhetorical theory but on performance-art theory. Rather than attempting to convey meaning or persuade via representation of arguments in processes, one ought rather to design games aimed at setting up meaningful situations or effecting interventions.
If I'm correct (or close?), that feels like a more productive conversation to have. My boring non-polemical view would be that we should have lots of kinds of games doing lots of kinds of interventions.
But in particular, I don't find it useful to identify procedurality with a specifically rhetorical approach. Certainly that is one thing procedures can be used to do, but a good portion of procedural installation work is precisely focused on setting up interventions, provocations, interactions, or even abuse of the audience. To take one deliberately abusive proceduralist work, the documentary-generator Terminal Time lets the audience choose viewpoints, and then generates a biased documentary aiming to flatter those viewpoints in a way that increasingly embarrasses the audience with extreme viewpoints —an anti-religious rationalist documentary may start out with material about how Galileo was persecuted, but end with a full-throated paean to the greatness of the Chinese invasion of Tibet and its deposition of the horrible theocratic Dalai Lama. This is not an abusive game per se, but does aim to (procedurally!) abuse its audience; compare also Paul Vanouse's game/installation Follower (1995).
Two final uncollected sets of issues.
As Sarah Roberts points out (quoted in a Simon Penny article), co-creation is a tricky business, since the artist or designer setting up the context or situation can give the illusion of sharing a kind of "choice", while at a meta-level he or she has spent considerable time thinking about what choices and situations are likely to result and has planned for those, too: "Interactive is not a bad word, but the illusion that goes along with it is of a kind of democracy, ... that the artist is sharing the power of choice with the viewer, when actually the artist has planned every option that can happen".
Finally, systems are more complex, inscrutable, and non-human-controlled than I think many of the examples on either side give them credit for. A key problem with both some of the art-game and serious-game writing and the above-cited Wilson & Sicart paper is that they tend to focus on fairly simple systems which can be viewed as more or less transparent, meaning they need not be considered seriously and independently except as extensions of a designer: either they encode the designer's meaning, or they're the instruments via which a designer dialogues with a player, generally posing no complexities or surprises of their own. But systems are more interesting and elusive than that, as Turing noted. Even an activity as simple as playing with river rocks in a stream derives a key part of its interesting strangeness from nonhuman systems: unlike playing with a static material like Legos, playing with river rocks is inherently a play with systems.