So, after taking on a host of Plantinga’s arguments, Richard Carrier Feser’s Five Proofs. I haven’t read the book, but I have read a number of Feser’s posts on the topic, and so I was interested in seeing what Carrier had to say about it. Of course, the problem with Carrier’s posts is that they are probably at least twice as long as necessary, which, I suppose, isn’t something that someone who calls himself “The Verbose Stoic” should comment on. However, I think it is fair to note that for the most part the posts spend far, far too much time insulting and making snarky comments about the people he’s commenting on, as it seems that he spends as much time trying to convince the reader that the people he’s commenting on are ignoramuses as he does trying to demonstrate that the arguments are wrong. As as I’ve said a number of times in the past, the problem with that sort of approach is that you had better be right, because if you are wrong then you look really, really bad.
Carrier, it seems to me, gets quite a bit wrong in this post.
I’m not going to deal with everything that was said in the post, instead hitting on a few ideas that strike my fancy. After all, I haven’t read the book and so can’t defend the full arguments. That being said, surely Carrier will hold to the basic notions of academic integrity and quote enough so that I won’t need to read it to understand what Feser is arguing, right?
Let me start with a rather odd constant gripe of Carrier’s, in that he criticizes Feser for the number of premises that he uses in his arguments:
Feser’s formalization of this argument appears around page 35. It has 49 premises. I **** you not.
I’m not sure why this should be seen as an issue. A large number of premises in an argument, especially if they are all directly outlined as such, would merely suggest a potentially complicated argument, which is certainly not something to be surprised or annoyed at. And outlining that many premises directly will make it more difficult to have hidden premises that Feser isn’t acknowledging. Does Carrier not like reading or something? Does he want his arguments as short soundbytes as opposed to full arguments? I don’t know, but it seems rather odd for Carrier to harp on the number of premises, and he does that in pretty much every point he addresses.
But let’s get into the arguments. The first one is at least one version of the “Unmoved Mover” argument, and the first thing Carrier does is to make a move that I’ve seen before — and probably talked about — and is, well, a pretty bad one, as he tries to demonstrate that a real nothing will invariably produce a something (as usual, Carrier italicizes too much for me to go back and fill it all in, but go look at his post to see his emphasis):
But what happens when you take away everything except that which is demonstrably logically necessary? Not what we “conjecture” or “wish” were logically necessary; no, we don’t get to cheat. No circular arguments. Only what we can actually formally prove is logically necessary. And that means, prove now, not at some hypothetical future time. We don’t get to “conjecture” or “wish” into existence some new logical necessity we have yet to really prove is such. Well. What happens is, we get a nothing-state that logically necessarily becomes a multiverse that will contain a universe that looks just like ours. To a probability infinitesimally near 100%. See Ex Nihilo Onus Merdae Fit.
A quick and dirty way to phrase that argument is: if nothing exists, then by definition no rules exist limiting what will happen to it; if no rules exist limiting what it will happen to it, it is equally likely it will become one of infinitely many arrays of things (including remaining nothing, which is just one of infinitely many other things no rule exists to prevent happening); if we select at random from the infinitely many arrays of things it can become (including the array that is an empty set, i.e. continuing to be nothing), the probability is infinitesimally near 100% the array chosen at random will be a vast multiverse whose probability of including a universe like ours is infinitesimally near 100%. Because there are infinitely more ways to get one of those at random, than to get, for example, the one single outcome of remaining nothing. There is no way to avoid this. Unless you insert some law, power, rule, or force that would stop it, or change the outcome to something not decided at random. But once you do that, you are no longer talking about nothing. You have added something. Which you have no reason to add. Other than your human desire that it be there. Which is not a compelling argument for it being there.
This is, to me, an argument that is so bad that it’s hard to know how to attack it. This is not a case where it seems like there’s something wrong with it but I can’t say what, like is often said about the Ontological Argument, but instead that there are so many things wrong with it that it’s hard to know where to start to express that clearly. Well, let’s start with this: the reason that an absolute nothing can’t produce anything is not because of some rule that says that it can’t happen, but because it lacks specific powers, namely all causitive and creative powers. There are no movers in absolute nothing, and no potential movers, and so no potential movement (referring back to the “Unmoved Mover” discussion mentioned above, which relates more to causes in general in today’s terms than mere movement in space). No potential movement, no possible way to change. Thus, nothing can ever change and so nothing will ever happen to it. Thus, no random events that will eventually likely produce a multiverse.
Carrier is going to try to argue that we still need a rule to say that, and so a rule has to exist, and so I wouldn’t be talking about nothing. I think he explicitly says that in the linked post in the quote above, that you have to have rules of logic to say that and rules of logic are things. But, the rules of logic don’t work that way. It is not the case that if I try to create something in this world that it is logically impossible to create, I can get part way through the process but at the point where the logical impossibility would kick in the law of non-contradiction intervenes and causes the attempt to fail. I just could never do it. For Carrier’s argument to work, the laws of logic would have to be things that have causal powers. And at least in how we use them they, in general, aren’t things that have causal power in and of themselves. We use them to, essentially, describe what is true and what is false about a given situation and set of entities. In the absolute nothing, there is nothing to describe. Whether we can say the laws of logic “exist” or not is tricky and plays terribly with our intuitions, but it is definitely true that there is nothing and no relations to describe, so their application to the absolute nothing would be meaningless. And so we’d return to the basic presumption we’d need for something to be considered an absolute nothing: there is nothing there that has causal powers, and no laws or relations and nothing to have relations with anything else there (because there’s nothing there). And if nothing has any causal power or any potential for change, then nothing can change. If Carrier wants to argue that we’d still have to have laws of logic and so wouldn’t have nothing, that’s fine, but even if we remove them we’d still have to maintain there there is no causation and so no possibility of change … and we’d have to doubt that any kind of absolute nothing of the sort Carrier describes — and uses to build out his multiverse theory — could ever work because the only reason we insist that the laws of logic exist in Feser’s absolute nothing is because we know about them and can apply them there. Carrier would need an absolute nothing where we couldn’t say anything about it … even that anything can happen in it and does happen in it at random, which is what he needs to make his case.
Carrier seems to miss the key point here: the absolute nothing has nothing that has any casual powers, and so nothing can happen in it. In order to get something from that nothing, he has to find something that has causal powers that exists in it, and there can’t be any such thing. And if he tries to invoke causation from outside of it — by claiming that there is another universe that triggers us to arise from “nothing”, for example — then we don’t have absolute nothing because that thing outside of that nothing exists and is not nothing. Feser doesn’t have that problem for God because he explicitly denies that we started from nothing. Carrier, on the other hand, is trying to start from nothing, and that causes him all sorts of problems.
Carrier tries to call out one of Feser’s premises as being a false dichotomy, but does so in a way that’s … suspicious, to say the the least:
Most of them are uncontroversial on some interpretation of the words he employs (that doesn’t mean they are credible on his chosen interpretation of those words, but I’ll charitably ignore that here), except one, Premise 41, where his whole argument breaks down and bites the dust: “the forms or patterns manifest in all the things [the substrate] causes…can exist either in the concrete way in which they exist in individual particular things, or in the abstract way in which they exist in the thoughts of an intellect.” This is a false dichotomy, otherwise known as a bifurcation fallacy. It’s simply not true that those are the only two options. And BTW, this Premise, is the same key premise (hereafter always hidden) in all five of his arguments. We can thus refute all of them, by simply refuting this single premise (more on that later).
So let’s do that.
Ironically, a third option that in fact I’m quite certain is actually true, is the very option described by Aristotle himself. Aristotle took Plato to task for the mistake Feser is making, pointing out that it is not necessary that potential patterns actually exist in some concrete or mental form. They only have to potentially exist. Hence Aristotle said of Plato’s “world of forms” what Laplace said to Napoleon of God: “Sir, I have no need of that hypothesis.” Potential things are by definition not actual. So obviously we don’t need them to be actualized to exist. That’s a self-contradictory request. It’s thus self-contradictory of Feser to insist that potential things must be “actualized” somewhere (a mind; concrete things). Obviously there is no logical sense in which they must be actualized in that way.
Aristotle argued that potentials exist inherently in everything, without anything further needing to be the case. A cube contains the potential to be a sphere (by physical transformation); but not as if that potential is some sort of magical fluid contained physically inside the cube. It’s simply a logically necessary property of any material that it can be reshaped; if it can have shape, it can have any shape. Period. It is logically necessarily always the case.
So, Richard Carrier found an argument in Aristotle that Feser, who primarily works in Aristotlean and Thomist philosophy, missed, and seemingly is obvious as it seems to be a major argument that Aristotle made against Platonic Forms. Um, yes, that’s entirely credible and I immediately grant that. Or, rather, not. If the argument is as obvious as Carrier makes it sound, then I certainly would expect Feser to have caught it. You can argue that this is an Argument from Authority, but I have to ask you what seems more reasonable: that Feser, a trained philosopher who specializes to a great degree on Aristotle missed such an obvious counter, or that Carrier, who is primarily a historian and is an amateur philosopher, missed something in Feser’s argument that would show why Aristotle’s counter doesn’t apply to his argument. And Carrier only provides a small snippet of Feser’s overall argument here, so I can’t even check to see what Feser might have said to dodge this counter … although one obvious immediate and likely idea is that the forms and patterns Feser talks about are not Platonic Forms, and Aristotle’s alternative only applies to Plato’s Forms. But that’s as far as I can get without reading Feser’s book or without Carrier quoting the context more and outlining the details of the argument. Without that, however, I’m not going to accept that Feser missed something so obvious just on Carrer’s say-so.
Carrier next tries to address the ultimate substratum, and then to propose an alternative to Feser’s ultimate substratum — space/time — without having the need for it to be intelligent and conscious:
So Feser is just arguing space-time is God. Mindless, valueless, merely physical space-time. That’s just atheism.
What this means is that Feser’s entire book is about a single maneuver: trying to dodge that outcome by trying to bootstrap space-time into being an intelligent consciousness. But that’s where his argument becomes 100% bullshit. In no way does the substrate having these other properties entail it’s “intelligent.” Intelligence is only a potential thing space-time can manifest, being an organized complexity; and being an organized complexity, it cannot be a property inherent in space-time itself, which is simple and uniform. Nor would it be “omniscient,” knowledge being another organized complexity, and thus only something that space-time can be organized to manifest, not a thing space-time itself is. All possible knowledge and all possible intellection is inherent in space-time as a potential, but that is not what we mean by knowledge and intelligence. Potentially knowing everything, is not the same as actually knowing everything. A clump of goo is potentially intelligent. Organize it into a functioning brain, and it will be actually intelligent. They are not the same thing. And “we” are indeed a way the universe becomes conscious of itself; but that does not make the universe a god. Not by any definition pertinent to anyone, least of all Feser.
Now, I do know quite a bit about the ultimate substratum, having talked about it before on a number of occasions. I’m pretty sure that Carrier doesn’t think such a thing is actually necessary, but is going along with it for the sake of argument. Fine, but if he does that then he needs to accept the reasons that Feser feels we need an ultimate substratum, and the reason he says we need it is because no property at this level can be actualized unless that property is actualized in the ultimate substratum. And so if it can’t be actualized at this level because it doesn’t exist in the ultimate substratum, then we can’t have it as a potential at this level either. Thus, a brain could have no potential for intelligence and so we could never have actualized intelligence in a functioning brain, no matter how we organized it. Thus, no inherent potential for intellection without it being actually actualized in space-time, thus making space-time intelligent. And if he has to add in all of those mental properties, then he ends up pretty much with a god.
That’s one of the most common mistakes people make in dealing with these sorts of arguments. Carrier presumes that Feser is starting from God and going from there to say what properties the ultimate substratum has to have, which is why he thinks he can get away with simply inventing something else that has those properties. But that’s not Feser’s argument. Feser is looking at what properties are actualized at this level and from there arguing to what properties the ultimate substratum has to have, and then saying that pretty much seems to be God. And if Carrier can in any way break the argument that the substratum doesn’t have to actualize the properties of this level, then he doesn’t need the space-time alternative to test, as he would have dealt the argument itself a major blow. So space-time does not do as much work as Carrier thinks it does. Which is a problem because he relies on that to do much of the work in the remaining points as well.
I’m going to skip the second argument as it mostly repeats the comments from the first argument, while focusing on “holding things together” which I can’t be bothered to get into, especially with Carrier’s sparse quotes on what Feser is actually arguing. I’ll also skip the third argument because that relies on Carrier’s odd idea of universals, which I’d rather get into when I talk about morality (which I still hope to do at some point). So now I’ll start talking about essences:
Even from a formal standpoint, this one is just a terrible mess. His syllogism has a ton of boner mistakes in it; for example, Feser’s Premise 2 (around page 128), asserts that “If [the distinction between an entity’s essence and its existence] were not a real distinction…then we could know whether or not a thing exists simply by knowing its essence.” Um. Yeah. That’s how we know dragons and unicorns don’t exist, and lions and tigers do. Because it would be impossible to know the complete essence of, say, a unicorn, and not notice that among its properties is the feature of “being fictional.”
Earlier, Carrier dismisses the idea of essences actually existing, which is not something that I’ll challenge here, especially since, again, due to a dearth of context I’m not sure how Feser is using “essence” here. But I will say that in this counter Carrier confuses essence as thing vs essence as set of essential properties that a thing must have. From later:
A fully informed account of an entity’s essence would include when it exists or didn’t. It is essential to Hitler, for example, that he did not live in the 21st century. It is essential to Yoda, for example, that no one could ever have spoken to him—other than in fiction or pretense. You could not fully understand what “Hitler” or “Yoda” were if you weren’t informed of these facts. And just excluding that one piece of information, literally the most important one, from what you will arbitrarily classify as “an essence,” is just a semantic game. And semantic games can’t get you to any grand realizations in metaphysics.
Feser actually burns a few pages arguing he is not engaging in this confusion. But alas, his protests make no logical sense. He insists if you mistakenly think lions are fictional monsters, “you have not misconceived what it is to be a lion.” Um. Yes. You have. You’ve totally misconceived what it is to be a lion. Only if you arbitrarily demarcate how you’d test whether a lion existed, with the outcome of that test—as if somehow the latter was not an attribute of the lion—can you get to Feser’s ridiculous premise. But that’s completely arbitrary. Why are we demarcating away that single property of lions as no longer essential to being a lion? Just because I know how to detect a dragon if one existed, does not mean I am necessarily fully informed as to what it is to be a dragon. If, unbeknownst to me, dragons exist, then I am simply misinformed about dragons.
In order to test whether dragons or unicorns exist, the first thing we need is a set of properties that would allow us to determine whether something is or is not a dragon or a unicorn. Ideally, we’d want the ideal set of properties that identify a dragon and only a dragon, the set that all dragons have in their entirety and nothing else does. Thus, we’d want to understand what it would mean to be a dragon completely and totally. If Carrier is right here, then that understanding would have to include whether or not it exists, as that’s part of the full understanding of what a dragon is. But then Feser’s counter comes into play: if we knew what its essence was, then we would always know whether or not it exists, and we can add on then that testing for it would be pointless. So if we need to test to see if it exists, then we have to test before we understand what its essential properties are, and if we wait to test it until we know what its essential properties are, then testing it is pointless. All of this points to the idea that whether or not something exists is not an essential property of the thing. And this is particularly true in this case, since what we are asking if the set of dragons or the set of unicorns is non-empty, and so we certainly need to know what would define that set before we can ask if it has any members.
Ironically, Carrier’s own space-time hypothesis works against him here, as it implies this:
Spacetime can be completely empty. And still have the potential to form up into matter, and thence a tree. In fact, it’s statistically inevitable that every bit of spacetime there is, will. Someday. It’s a Boltzmann necessity.
Essentially, the idea is that since space-time can and does change at random, eventually some part of space-time will form a complete brain without a body. Carrier seems to extend that to everything here, as that is the implication that allows for Boltzmann Brains. But then this means that at some point in time in the universe that a dragon will exist, or a unicorn will exist, for at least a brief period of time. And if that’s the case, then it could in fact be that way right now. Add in parallel evolution on the infinite number of planets that exist in the universe and the chances of a dragon or a unicorn existing somewhere in the universe are pretty good, much better than the 0% that Carrier needs to make his point here. And while with Hitler or Yoda he could avoid this by talking about a specific case, here he can’t, because we aren’t talking about specific dragons or unicorns, but instead about instantiations of the overall category, and so if any dragon or unicorn exists anywhere that would, according to Carrier, change its essential nature.
On top of all of that, this works out badly with the potential/actual model that Carrier is supposedly sticking to for the sake of argument here, because if you change the essential properties of an object then it by definition is a different object. Thus, an object that comes into existence does not merely actualize its potential, but instead changes into a completely different object. That’s not how I see actualization of a potential, even that of coming into existence, to work, and so that his move here makes that model problematic suggests that, again, he’d have to abandon that model for his argument, which would be far more serious an issue.
The last argument relies on the space-time alternative that I’ve already shown problematic, so this is a good place to stop.
I’m not saying that Feser’s arguments work; in general, I don’t buy those sorts of arguments myself. I’m not saying that Carrier hasn’t found problems with Feser’s arguments; the context is so vague that I’m not really sure what Feser’s argument really is most of the time. What I’m saying here is that a lot of the key arguments Carrier relies on are … not good, to say the least, and are not good in ways that really, really bug me. Hence, the post pointing out those not good arguments and showing why they are not good.
Extra Credits on the Cost of Making Games …
February 28, 2018So, as a follow-up to the video talking about how much video games should cost, Extra Credits did a video talking about how much it costs to make a AAA game and about how there isn’t really that much room to cut costs there. I’m not going to get into whether, for example, marketing is as important as they make it out to be or whether graphical fidelity is really as important to the console market as they say, but what’s interesting about it is that they take the idea of as slimmed down a studio as you can get and say that that would be 200 people working for 2 years to finish a AAA game. It so happens that I, in fact, worked in a 200 person department making software (not games), and I can say with certainty that if that’s the minimum case it isn’t at all sustainable.
The issue is that those 200 people would be working on a product for two years, and the company would see little to no revenue from that product for that entire two years. In our case, we were releasing an update to the product at least once a year if not multiple times a year, and also were adding things to that product and not starting over roughly from scratch. While franchise games aren’t quite starting over from scratch, even they are doing more than adding new features to a pretty much finished existing game. The closest we have to that are the sports games that EA does so well, and that’s only if they only do roster changes and not engine updates or major features. In our model, if we wanted a completely new product — particularly if that was supposed to be a replacement for the existing product — what we did was use the old product’s revenue to fund the new product. We’d shift people away from the old product to work on the new one, while still maintaining enough people to add new features and fix bugs to generate new releases that generated enough profit to cover the development costs of the new one until it had enough market share and revenues to generate profit on its own. But we could do that because we were developing mission critical software that our customers did not want changed. They didn’t want new interfaces or new engines or any kind of new experience. New experiences with that software cost them money in terms of training and potentially in the effectiveness of their operators. They wanted, ideally, one product that did everything they wanted it to do and that when they needed new technologies it was just added to this product and worked in pretty much the same way as all the other technologies did. That’s obviously not true for games, where customers generally purchase a number of widely varying products that they can roughly use in parallel.
So, what we have in the gaming space is a product that takes two years to make a game, costs a fortune to make — again, a 200 person department/studio has a large number of fixed costs — and the game has to earn its money back in a shockingly short amount of time, because games don’t have a huge productive shelf-life. You’re looking at a percentage of revenue after one or two years even if you gradually reduce prices that the product I was on had during the last 2 – 5 years of a 20+ year run. And you have to be working on the next game almost immediately, even before you see how well this one has done, because you can’t stretch the next game out for longer than two years. So we can see why AAA game developers would really like to get a model like ours was, where they build a basic software product and do add-ons to it for the next while, so instead of developing an entire new game over two years, they instead build a set of add-ons every six months, say, and get either a constant revenue per month (a subscription model) — or an in-flux of revenue when each add-on is released (DLC, expansions, and loot boxes). The EA services model and, of course, loot boxes are the more recent and controversial attempts to do just that.
The problem is that this model does not work very well for video games, at least in the long term. The customers for games generally don’t want to play just one or a small number of games for years and years. They like some variety, and to even have a stable of games that they play and so don’t even necessarily play the same game the next day. They also have strict time and budget restrictions when it comes to playing games, since it comes out of their entertainment budget. The more that they play a specific game, and the more they have to pay to keep playing that specific game, the less time and money they have to play or purchase other games. If one game becomes really successful and charges a significant enough percentage of the average gamer’s gaming budget, it can lock out other games as there just aren’t enough gamers left who are willing to play and pay for those other games while they spend most of their time on that one. This can not only hurt existing games, but can of course greatly impede new games, who can’t even break into the market because everyone else is still playing that old game. And this, of course, will also impact new games from the same studio, as their players are still playing the old game. Should the old game hit saturation where they can’t add anything interesting and their customers are tired of the existing game, they may — and are likely to — lose customers to something else faster than they can get an alternative up and running. If anyone still has any alternatives available.
And this isn’t actually speculation, but is something that we’ve seen already in the MMO space. In 2014 I talked about MMO saturation, with there being a large number of MMORPGs still running, with World of Warcraft being the killer game that hampered new MMOs when they came out. At the time, MMOs were the big thing, in large part because they provided this sort of revenue stream. But once WOW gained prominence, new games struggled, and a number of promising candidates faded away far earlier than the original games had, and weren’t profitable. Most MMOs had to adopt some kind of Free-To-Play model to generate sufficient revenue to keep going. While you can talk about the MMO shooter games like Destiny and Overwatch, MMOs aren’t the growth industry that they were back then, and it very much seems like they are fading a bit. I don’t know of too many new ones coming out, and Shamus Young — my main source for industry information — used to play a lot of them a lot.
If AAA gaming companies try for this expanded revenue stream, they are likely to run into the same issues. Even the sports sim genre hits this because the revenues from the previous years’ games tend to fall off fairly sharply when a new release comes out, but this steady revenue system wants to prolong the life of existing games. There are some ways to mitigate this — extra things being compatible with multiple games, for example — but aiming for this sort of revenue stream is not going to help games as much as you might think because of the very nature of games. If you prolong the life of a game, people don’t buy new ones, but you need them buying new ones, too.
Posted in Not-So-Casual Commentary, Video Games | 1 Comment »