I’m currently working on a paper that focuses and refines my examination of a question posed in a 2016 blog post called “Free Will Paradox?”
Related to that question is an idea I’d like to explore further but that doesn’t fit into the new paper. So I’ll do it here, borrowing freely from the above post.
From Determinism to Randomness to (No?) Free Will
Contemporary discussions of free will usually follow this path:
(1) Discuss the (worrisome or not) implications of determinism for free will. Where “determinism” generally means that the state of the world at this moment is determined, or “caused,” by the state of the world at the moment just before this one, and so on back to the Big Bang.
(2) Observe that, according to dominant interpretations of quantum mechanics, we live not in a deterministic world, but in an inherently random one. Where “inherently” implies an irreducible, objective, or “quantum” randomness, in contrast to “so-called randomness” arising merely from ignorance.
In other words, some human decisions and behaviors may be influenced by inherently random (or “indeterministic”) forces, rendering determinism false.
(3) Conclude that quantum randomness isn’t any more reassuring for free will than determinism is.
I’d like to present some alternatives to (3), as I reject the conclusion that those who believe in free will need to worry about randomness.
A typical way to justify this rejection is to point out that randomness is only a micro concern, while for us in the macro world, we might as well assume a Newtonian, or deterministic, model in which randomness is epistemic (i.e., is due to lack of information).
That makes sense to me. But I’d like to reject the randomness worry on other grounds that I haven’t seen before (if you have, let me know where).
First, I’ll say a little more about where the debate is on the randomness question by looking at a few high-profile examples. These include typical and not so typical rejections of (3).
Views of Quantum Randomness and Free Will
I haven’t seen a persuasive account—or even attempts at an account, really—of how quantum randomness could possibly affect the goings-on of comparatively large brain structures in such a way that directs my choosing between, say, a peach and cake slice. (The example is borrowed from Thomas Nagel’s lucid “Free Will” chapter in his 1987 book What Does It All Mean?).
A technical expression of why this might be comes from physicist Max Tegmark in his 1999 paper “Importance of Quantum Decoherence in Brain Processes” (Physical Review E, Vol. 61, Iss. 4 , 4/2000; or read the full article here). The abstract:
Based on a calculation of neural decoherence rates, we argue that the degrees of freedom of the human brain that relate to cognitive processes should be thought of as a classical rather than quantum system, i.e., that there is nothing fundamentally wrong with the current classical approach to neural network simulations. We find that the decoherence time scales (~10-13 − 10-20s), are typically much shorter than the relevant dynamical time scales (~10-3 − 10-1s), both for regular neuron firing and for kinklike polarization excitations in microtubules. This conclusion disagrees with suggestions by Penrose and others that the brain acts as a quantum computer, and that quantum coherence is related to consciousness in a fundamental way.
The idea here is that quantum states are too short lived in the microtubules (a little tube inside neurons) to contribute to human consciousness. This seems to also have clear implications for the ability of quantum states to influence the direction, so to speak, of consciousness with respect to decision making.
As Tegmark is puts it in an interview with The Guardian that same year:
…we would all like to believe that something amazing is happening in our brains. However, Tegmark thinks whatever is going on is still amazing.
“It’s just that it does not involve quantum processes in any important way,” he says.
[Source: “How the Mind Takes a Quantum Leap” (10/20/99), by Marcus Chown.]
Plenty of people agree with Tegmark. Science and mathematics reporter Charles Seife, for example, seems persuaded in his Science article, “Cold Numbers Unmake the Quantum Mind” (2000), given its summary, which states that Tegmark “has concluded that possible quantum microprocessors decohere far too rapidly to orchestrate the firings of neurons.”
More recently, without reference to Tegmark’s calculations, the same basic conclusion is implied by physicist Ian Durham in a presentation shared on the FQXi Podcast episode “Measuring Free Will” (9/12/19 in my app’s feed, or watch at YouTube under the title “Ian Durham – A Formal Model of Consciousness and Emergent Free Will“). As I understand Druham, quantum randomness may be practically ignored when modeling macro-systems. For example:
…we were asked about how you could get something… our… sort of deterministic world out of something more random, right? Underlying randomness. And it turns out this is actually a pretty standard thing that happens in statistical mechanics, for instance. So if you have interacting systems of random processes, they can actually converge to a deterministic macro-state if the multiplicity is high enough. This happens all the time, in fact. This is called the thermodynamic limit. (starts at 7:05 on YouTube video)
The focus of this presentation is Durham’s work on a new mathematical model of free will, about which he says a paper is forthcoming. The model he describes, with appropriate modesty, sounds interesting (he says it is panpsychist and that he’d like it to be compatible with Information Integration Theory). In the meanwhile, check out the above podcast and/or this accompanying blog post (8/14/19) he wrote, which is worth excerpting to get a sense of his conception of free will and, just as importantly, his optimism:
So one possible measure of “free choice” would be some function (which I call the zeta-function) of the minimum Mahalanobis distance between that choice and any other choice in the ensemble. Likewise it would be proportional to a time function that ensures that the full ensemble of choices can be read into our memory and processed in a finite amount of time. The zeta-function should also be inversely proportional to the variance of the choice (the smaller the variance, the greater the freedom as I just described).
The zeta-function is a measure of the freedom of an individual choice. It is safe to say that sometimes our choices are not free. Many actions in our life are instinctual. But it’s safe to say that if a certain number of our choices are free we might say we have free will. So a measure of free will itself, which I call the Z-function, would be some function of the collection of zeta-functions for all the choices we have.
Of course, one need not say anything about brains or free will in order to weigh in on views about quantum randomness. Here’s physicist and statistical mechanics/probability expert E.T. Jaynes, in Chapter 10 (section 10.7 “But what about quantum theory?”) of his intimidating (and posthumously published) 2003 text, Probability Theory: The Logic of Science:
Those who cling to a belief in the existence of “physical probabilities” may react to the above arguments by pointing to quantum theory, in which physical probabilities appear to express the most fundamental laws of physics. Therefore let us explain why this is another case of circular reasoning. We need to understand that present quantum theory uses entirely different standards of logic than does the rest of science.
In quantum theory …. when no cause is apparent one simply postulates that no cause exists—ergo, the laws of physics are indeterministic and can be expressed only in probability form. … The mathematical formalism of present quantum theory—incomplete in the same way that our present knowledge is incomplete—does not even provide the vocabulary in which one could ask a question about the real cause of an event. …
Quantum physicists have only probability laws because for two generations we have been indoctrinated not to believe in causes—and so we have stopped looking for them. (pp 327–328).
I take it Jaynes’s view is roughly in line with Albert Einstein’s “hidden-variable” rejection of the irreducible randomness of quantum events.
I’ve also encountered dismissal of such influences from non-physicists. Such as from independent scholar and probability expert Nassim Taleb, when he writes in his influential book, The Black Swan (make sure you read the 2010 second edition!):
If there is one thing on this planet that is not so uncertain, it is the behavior of a collection of subatomic particles! Why? Because, as I have said earlier [i.e., about why his coffee cup doesn’t spontaneously jump], when you look at an object, composed of a collection of particles, the fluctuations of the particles tend to balance out. (p 287)
Of course, plenty of people of brilliant people believe that quantum mechanics matters in the macro world, including at the level of neurons to such a degree that influences cognitive states (and thus behavior). Tegmark’s paper is in response to work developed by physicist Roger Penrose and anesthesiologist Stewart Hameroff, the latter of whom responded to Tegmark in collaboration with physicists Scott Hagan and Jack Tuszynski, saying why they believe Tegmark is wrong.
One criticism is that Tegmark’s calculation may apply to most of the brain, a generally wet environment, but not the microtubules which insulate against that environment. A starting place for more on this is Hameroff’s Wikipedia entry, a fascinating entryway to the multidisciplinary culture of consciousness studies in general.
I don’t know what their views are on free will in light of quantum randomness. The usual view for some time now, I think, is that one’s actions being random rather than causally determined doesn’t do any favors for the idea that one’s choices are under the control of one’s freely willed agency.
Physicist Adam Brown agrees, but on the surprising grounds that “in some sense quantum mechanics makes the behavior of agents not less predictable but more predictable,” as he puts it in another FQXi Podcast episode, “Quantum Mind Reading” (9/7/19 in my feed; or watch it on YouTube, where the title is “Adam Brown – Agency and Quantum Mechanics,” and the quote starts at about 1:20).
Brown doesn’t give a justification here for the premise that predictable behavior is unfree. Though, to be fair, it’s a commonly presumed premise, such as in discussions of Libet-type studies. There are various arguments against it, however, my favorite of which I explore in my post, “Laplace’s Demon Defeated by Human Consciousness.”
At any rate, Brown expresses his free will skepticism in contrast to a quote that he shares from physicist Arthur Eddington, who was, as far as I know, the first to proclaim anything about the significance for quantum mechanics on free will. I’ll share more of it than Brown does, as there are some subtleties in it that go beyond Brown’s discussion. It comes from Eddington’s 1928 book, The Nature of the Physical World, Chapter XIV, “Causation” (after noting that :
In the old conflict between freewill and predestination it has seemed hitherto that physics comes down heavily on the side of predestination. …
On the scientific side a new situation has arisen. It is a consequence of the advent of the quantum theory that physics is no longer pledged to a scheme of deterministic law. …
The future is a combination of the causal influences of the past together with unpredictable elements—unpredictable not merely because it is impracticable to obtain the data of prediction, but because no data connected causally with our experience exist. It will be necessary to defend so remarkable a change of opinion at some length. Meanwhile we may note that science thereby withdraws its moral opposition to freewill. Those who maintain a deterministic theory of mental activity must do so as the outcome of their study of the mind itself and not with the idea that they are thereby making it more conformable with our experimental knowledge of the laws of inorganic nature. (pp 293–295)
The subtly is in Eddington’s injunction to study the mind itself. Without trying to read too much into what he means by this, I will presume that the condition is roughly satisfied by a brain researcher who takes the view that randomness in the brain supports the existence of free will.
Neurobiologist Martin Heisenberg is who I have in mind. I’m going to linger on this one a moment.
The randomness in question—which Heisenberg generally means to be of the irreducible, quantum sort, though he maintains that any sort of randomness will do*—shows up in, for example, “the random opening and closing of ion channels in the neuronal membrane, or the miniature potentials of randomly discharging synaptic vesicles,” to quote Heisenberg in his 2009 Nature article, “Is Free Will an Illusion?” (5/13/09).
[*When asked about this in a conference discussion, he reminds us that he is not a physicists, but a biologist; watch on YouTube: “The Role of Objective Chance in the Brain and Behavior,” about 35 seconds in. It strikes me that if the randomness is epistemic rather than irreducible, then the processes he describes are determined rather than “self-generated” (more on this below). And so perhaps this does bring us back to a starting point to do with questions about physics. The point I’ll pursue here, though, is about the philosophical interpretation of the case in which the randomness is indeed irreducible.
(One might make a case for self-generation that looks like this: epistemically random processes being initiated by inner biological conditions that do not rely on input from external stimuli. I won’t get into this here, but will say that it will be difficult to clearly disentangle “external” and “internal,” as well as what the literal, rather than conceptual, boundaries are for the events in question. The long and short of it, I think, is that Heisenberg’s picture requires, within the organisms biology, some point of irreducible randomness between the birth of the organism and the action in question.]
The idea here is that we have free will because the factors that influence behavior are not strictly due to external stimuli, and instead are due to goings-on in one’s own biology.
In other words, Heisenberg’s model relies on a definition of “free will” in which “we need not be conscious of our decision-making to be free. What matters is that our actions are self-generated.” The idea occurred to him while studying fruit flies, which exhibit self-generated behavior. He also notes that “randomly generated action” that “does not depend upon external stimuli” is found in the the “way the bacterium Escherichia coli moves.” I suppose this means they may have some degree of free will.
This picture of free will doesn’t capture what many, skeptic or believer, consider to be free will’s most important feature: a necessary condition for considering ourselves morally blameworthy for our choices and actions. Usually, unfettered free will makes us responsible for our choices and actions, but Heisenberg’s free will seems to make us less responsible.
As Sam Harris put it in his response to Heisenberg: “If I were to learn that my decision to have a third cup of coffee this morning was due to a random release of neurotransmitters, how could the indeterminacy of the initiating event count as the free exercise of my will?” (from The Moral Landscape [2011], p 104).
I don’t see how this picture could result in deciding to have a third cup of coffee. I have not no account of how any of these events amount to your brain either putting the idea of a third cup of coffee into Harris’s head, nor, alternatively, how these tiny brain parts could notice that Harris is deliberating about having a third cup, and then somehow flip a coin for him while harboring the representation of a sample space of “{(third cup), (no third cup)}. In the next section, I’ll say more about this sort of worry.
I primarily bring up Heisenberg’s example to demonstrate that people talking about free will may be talking about quite different things, sometimes things that don’t seem at all like what most people mean by “free will.” (Indeed, most philosophers and scientists having this debate, believers and skeptics, agree that hardcore libertarian free will, of the sort a Christian or even an everyday “non-expert” might care about, is impossible.)
The example also offers a glimpse of how many views there are on the question. I could spend a lifetime enumerating them. If that sounds like fun, a good place to start is the Information Philosopher website, perhaps with this entry: “Martin Heisenberg.”
(Quantum) Randomness Doesn’t Matter to Free Will
My intuition is that those arguing against quantum randomness influencing human agency are correct. But from here on, I’m going to assume that they are wrong.
My claim is that randomness is not a problem for free will even if it does affect the macro world, but not in the way Heisenberg supposes.
Let’s suppose that tiny random events, quantum or otherwise, influence neurons enough to influence human agency. My bet is that this would, at most, result in performing an activity in a slightly different manner than you would have otherwise. But it wouldn’t result in performing a different action entirely.
In other words, randomness will not decide whether or not you choose the peach, but it might result in your picking the fruit up slightly more quickly or slowly.
That’s my bet, because it seems to me that the alternative influence would result in widespread, chaotic behavior that we simply don’t observe.
Here’s what I mean. Suppose that randomness does influence which action you resolve to perform. I don’t see why the chosen action would need to have anything to do with the options between which you are deliberating.
While trying to choose between the peach and cake, it seems arbitrary to assume that randomness would lead to one of those specific activities, rather than, say, spinning in a circle two times or blurting the words “Anti-Skeptic Spray.”
Quantum-scale entities don’t have the relevant representations—or any representations at all—that would allow them to construct the complex collection of brain processes tantamount to resolving to eat the peach, reaching out and grabbing the peach, buying it, and so on. It’s not as if there’s a neuron that will, depending on some quantum outcome, fire left or right, or fire or not fire, such that the brain state for “I choose the peach” is formed. Now apply this to many many neurons at once.
(Maybe such random activity is going on, but it simply results in noise that at most could be distracting, but more likely would amount to nothing experienced by the chooser.)
The states involved in all this involve relations between brain matter. They have no meaning for the neurons or their constitutive parts. Nor does the idea of “choosing” or “deliberating” have any meaning for these relations that happen to correlate with, or be the basis for, experience of a certain kind (of the kind that feels like “deciding” or “being torn”) in the chooser. These relations are bound up in the intractable problems of consciousness, of relating neurons to experience, of mental causation, and so on.
This, by the way, is in line with the reasoning often given by proponents of the many-worlds interpretation of quantum mechanics for why it is wrong to say that, every time we make a decision, a new world is created. For example, from physicist Sean Carroll’s 2019 book, Something Deeply Hidden:
…No, you do not cause the wave function to branch by making decisions. In large part that’s just due to what we mean (or ought to mean) by something “causing” something else. Branching is the result of a microscopic process amplified to macroscopic scales: a system in a quantum superposition becomes entangled with a larger system, which then becomes entangled with the environment, leading to decoherence. A decision, on the other hand, is a purely macroscopic phenomenon. There are no decisions being made by the electrons and atoms inside your brain: they’re just obeying the laws of physics.
Decisions and choices and their consequences are useful concepts when we are talking about things at the macroscopic, human-sized level. It’s perfectly ok to think of choices as really existing and having influences, as long as we confine such talk to the regime in which they apply. … Your decisions don’t cause the wave function to branch, because “the wave function branching” is a relevant concept at the fundamental level of physics, and “your decisions” is a relevant concept at the everyday macroscopic level of people.
So there is no sense in which your decisions cause branching. … even one that seems like a close call at the time. (pp 214–215)
Carroll then says a few words about the mechanics and chemistry of neuron interaction, noting that “for the most part, these processes can be thought of as being purely classical, or at least deterministic. … So ‘decisions’ are best thought of as classical events, not quantum ones.”
I take it that we can conceive of “decisions” as macro events whether or not when quantum-scale entities go about “just obeying the laws of physics” that means they dissipate too quickly to exert influence (as Tegmark suggested) or not. I’m considering the “or not” case, while pointing out that there is no way in which these minuscule goings-on have any representation of “peach vs. cake.”
I arrived at this thought before encountering such an explanation as Carroll’s, as it follows just as readily from contemplation of philosophy of mind. In fact, what I’m saying is more involved than the above, perhaps goes a bit further. Quantum-scale entities have no stake at all in what you do, no sense of the meaning behind the objects of your deliberation, no phenomenology or representation of them or of anything else. They are involved in building representations, of building a mind, whose content consists of associations that lead to more such building, an interaction we don’t understand, but relies far too much on representational content—i.e., in the mind of the chooser—for that content itself to be influenced by things that have no content (i.e., by the the things on which that content supervenes; or from which it emerges).
Put another way, mental states must have some influence here that is stronger and further reaching than that of mindless sub-atomic particles, when it comes to the construction of new brain states on which supervene new mental content—i.e., in going from “peach or cake?” to “I’ll have the peach.”
I’ll try to say this in a less convoluted, more illustrative way, though by being so specific the example will become cruder than what I’m actually struggling to describe.
Suppose you choose the peach. To enact that choice requires a series of acts, such as grasping the peach. What you chose, however, was to eat the peach, or some similar macro action. You did not explicitly choose to, say, move your arm in the direction of a certain object (the peach), then wrap your fingers around that object, and so on.’
We could break this down into steps as small as we like: “I first chose to rotate my shoulder 1/1000000 of an inch, and then chose to rotate it another 1/1000000, and then another 1/1000000, and so on until the peach is consumed, barring some macro intervention; a quantum flicker isn’t enough.
The sum of this accumulation of activities is, “choosing to eat the peach.” Much of it, if we include the steps that follow the decision—the chewing, swallowing, digesting—is as automatic as playing a scale at the piano, riding a bicycle, or catching a basketball. (I think we must include at least an attempt or wish to execute those steps: were I to say, “I chose to eat the peach but ate the cake on my own volition, with no external intervention on my choice,” you would likely say, “no, you chose to eat the cake.”)
A quantum flicker would have no stake in that sum we call “choosing,” no sense of the meaning behind “peach vs. cake” even if it did have some influence on how one of the many given steps plays out. But initiating the entire sum is too meaningful for the quantum flicker.
There is no peach button and cake button towards which a quantum dart is blindly tossed by a homunculus in your head, such that whichever button is hit initiates the mental activities that amount to choosing peach or cake. Such a thing might, rather, influence some small step in how a made choice is enacted. Or, alternatively, if there are such buttons, the dart may hit any number of other buttons, having nothing to do with peach or cake (there’s no specified “dart board” that the homunculus would recognize and aim for).
I’ll now consider the point sufficiently belabored. And will note that I could easily speculate about some counter-explanation, but my efforts strike me as highly fanciful and overly specific. And probably untestable.
For example, suppose that, in deliberating, you develop two counterfactual mental models (an activity that is itself subject to quantum randomness, but never mind). These models have neural bases (or correlates) in the brain: one for choose to eat the peach, one for choose to eat the cake. You develop these models, and their neural bases/correlates, by imagining what it would feel like to eat the peach and the cake, and what consequences these choices will have, and so on.
These distinct models may be categorized as “competing activities” because they attempt to utilize related resources in the brain, or to directly satisfy some particular drive, and so on (it’s not like trying to decide between, “should I eat the peach or should I buy the house on 7th Ave?”).
Whichever of these structure is most robustly physically constituted—i.e., the first of them to result in a sufficiently explicit activity-inducing structure—wins. For example, “choose cake” surpasses the “executive” defenses attempting to veto breaking your diet, and initiates the decisive pro-cake move, which may at this point be executed with near-automaticity; almost like a sneeze. (Notice here that the drives being satisfied are overlapping, rather than being identical. Choosing the peach, for instance, may satisfy a desire to save money or to eat healthier.)
It could be that, if resources in the brain are being dedicated to allotting resources to “peach” or “cake,” a random event could tilt the balance just enough for one of them to win—i.e., for the brain states correlated with “choose peach” or “choose cake” to have the most constitutive brain-stuff accumulated. Why not? I have no idea. And it has elements that seem to align with promising work I’ve read about in embodied cognition.
What I don’t like about it is that it’s bullshit. I’m just making stuff up that feels sort of right based on years of reading and thinking and reading about this and related topics. These discussions tend to rely on a lot of assumptions and on metaphors that usually go unexamined (e.g., what, if anything, is a “brain state” anyway?).
That some, some of the people I’ve referenced here have developed some pretty sophisticated ideas (e.g., see the specific theories from Hameroff and collaborators for the neural basis for consciousness). My strong impression, however, is that most people who think randomness, quantum or otherwise, matters in the brain couldn’t give even a slight account of how such a thing might direct decisions. Generally, it seems to be taken for granted that, somehow or another, quantum randomness must be doing something.
The same goes, frankly, for how one brain state “causes” another more broadly speaking—which falls under the problem of mental causation (a term that some folks seem to think is too meaningless to utter).
All this vagueness, lack of explanations, and taking-for-granted is why I’m a free will skeptic on philosophical rather than scientific grounds (as I point out in my post “Free Will Skepticism Doesn’t Need Science”). In fact, I’d say that most of the claims I’ve seen about these things come down to being philosophical claims, no matter who’s making it—that is, when someone claims a priori that free will skepticism (or belief) is the sounder scientific position to take.
To say more about this would require discussion of an idea about the mind-brain relation I often put as, “the brain is not a Rube Goldberg machine.” I take the randomness-meets-neurons question to be a subset of this larger question. About which I’ll say only the following.
The Brain Is Not a Rube Goldberg Machine
I don’t see how the chain of thoughts kicked of by the thought “I’d like some ice cream”—which leads to associated thoughts about where to get ice cream, whether it’s too late to have sugar, what flavor I’d like, will I have to share it with Tony, etc.—is identical to neurons interacting like so many billiard balls knocking into one another. There must be a more complex (physicalist) story—perhaps a wholly different story entirely—to be told here than anything like what goes on with networks of interacting billiard balls (or toppling dominoes, etc.). Especially if the arrangement and activity of those balls at a given instant result in, or are identical to, a mental event.
For example, when I ask for your associations for the world “apple,” there are loads of things you might mention (first sin, The Beatles, granny smith, Steve Jobs), but those things don’t come to mind because a formation of neurons is knocking into other neurons, causing them to take on formations that correspond to those associations. While there is surely a physical story to tell that’s vaguely like that, it must also be true that the mental content “apple” (for an English-speaker of a certain cultural background in the year 2020) itself has influence as well—i.e., it isn’t merely an emergent epiphenomenon.
I think more folks—including neuroscientists, physicists, and computer scientists (though maybe not enough of the latter!)—are expressing agreement with anti-epiphenomenalism talk in recent years. But they also still talk as though we’ve got billiard balls in our heads—as though there is a contained system in which brain parts are mindlessly knocking into each other, subject only to the laws of physics with no intervention from mental or intentional content, and that this is the root of the aforementioned associations, rather than the activity of neurons somehow being influenced by the mental/representational content of those associations.
In other words, I’m claiming that, when you picture an apple in your mind and then spontaneously picture Steve Jobs, that association somehow arises in part because you were picturing an apple. Not just because the brain stuff on which picturing an apply supervenes knocked like so many dominoes into some other brain stuff that then by some bizarre coincidence was jostled into falling into a formation from which there emerges imagery of Steve Jobs. I think many people would reject that the way they talk implies anything so facile as this, but it seems to me that this is just what it implies, even if they attach a disclaimer that their talk can’t capture the true complexity of what is (or is assumed to be) going on. (I resist the temptation to cite examples today.)
To be clear, I’m not claiming that there is an escape from physical laws (physical law is still in play when you grab a literal billiard ball—when you enter into the “closed system”—in order to intervene on its otherwise mindless trajectory). Rather, I’m claiming that the talk I hear doesn’t make sense, is vague, relies on metaphors and assumptions, and is just all around sloppy and unsatisfying.
I can think of a constructive theory or two that may be exceptions to this. I’m not persuaded by them, but they are careful, at least. I’ll save those for another day.
Enjoy or find this post useful? Please consider pitching in a dollar or three to help me do a better job of populating this website with worthwhile words and music. Let me know what you'd like to see more of while you're at it. Transaction handled by PayPal.Or click the banner to shop at Amazon (at no extra cost: it just gives me some of what would have gone to Amazon).