Dr. Ewert answers Andy

Some of you are raring to go with responses to the Uncommon Descent post Dr. Ewert Answers. Adapa dropped the following into a thread in which I hope to engage Ewert in discussion of my own question, not Andy’s. Have fun.

___________________

Over at UD Ewert hand-waved away this question by Andy:

Andy: “At ENV you said: “The universe must have begun with a large amount of active information with respect to the target of birds” How did you determine birds were a “target”? What of the billions of other living and extinct species?

…by just C&Ping more of the orifinal ENV post.

Ewert: “One might ask, why birds? Birds are, in this discussion, thus far the target. By “target,” we do not mean something for which the search is actively looking. Recall that the only requirement of the search is that it be representable as a probability distribution. The target plays no role in what constitutes a search; rather, the target only features in the context of measuring the active information in a search. The target is effectively the measuring stick. The choice of target is arbitrary, and I could have as easily chosen cities, paintings, beetles, cows, volcanoes, mountains, lakes, or crystals. The same conclusion applies to all of them: they show up far more often than chance would lead us to expect. Within a materialist framework, they have to be explained by a process biased in favour of producing them.”

If any target will do then what’s the point of this “active information” bafflegab? Isn’t this just the lottery fallacy again?

As for the last part

Ewert: “The same conclusion applies to all of them: they show up far more often than chance would lead us to expect. Within a materialist framework, they have to be explained by a process biased in favour of producing them.”

No shit. The process is called evolution. Scientists have been studying it for 155 years.

15 thoughts on “Dr. Ewert answers Andy

  1. Over at UD Ewert hand-waved away this question by Andy:

    Andy: “At ENV you said: “The universe must have begun with a large amount of active information with respect to the target of birds” How did you determine birds were a “target”? What of the billions of other living and extinct species?

    …by just C&Ping more of the orifinal ENV post.

    Ewert: “One might ask, why birds? Birds are, in this discussion, thus far the target. By “target,” we do not mean something for which the search is actively looking. Recall that the only requirement of the search is that it be representable as a probability distribution. The target plays no role in what constitutes a search; rather, the target only features in the context of measuring the active information in a search. The target is effectively the measuring stick. The choice of target is arbitrary, and I could have as easily chosen cities, paintings, beetles, cows, volcanoes, mountains, lakes, or crystals. The same conclusion applies to all of them: they show up far more often than chance would lead us to expect. Within a materialist framework, they have to be explained by a process biased in favour of producing them.”

    If any target will do then what’s the point of this “active information” bafflegab? Isn’t this just the lottery fallacy again?

    As for the last part

    Ewert: “The same conclusion applies to all of them: they show up far more often than chance would lead us to expect. Within a materialist framework, they have to be explained by a process biased in favour of producing them.”

    No shit. The process is called evolution. Scientists have been studying it for 155 years.

  2. “show up more often that chance would lead us to expect” is just the eleP(T|H)ant again.

    Any process with a non-flat probability distribution is going to be “biased” in favour of some outcomes more than others.

    Is Active Information simply the non-flatness of the probability distribution?

  3. I commented in the other thread:

    Elizabeth:
    “show up more often that chance would lead us to expect” is just the eleP(T|H)ant again.

    Any process with a non-flat probability distribution is going to be “biased” in favour of some outcomes more than others.

    Is Active Information simply the non-flatness of the probability distribution?

  4. Adapa: If any target will do then what’s the point of this “active information” bafflegab? Isn’t this just the lottery fallacy again?

    Or rather the Texas sharpshooter fallacy. Whatever happens to be there “shows up more often than chance would lead us to expect”, just like that particular cluster of holes in the side of the barn.

  5. Elizabeth,

    Keiths previously created a thread for fifthmonarchyman, who wanted to respond to the question I’d posed to Ewert. I thought that Adapa’s comment was worthwhile, in and of itself, but irrelevant to my question of model reification. Now you and Piotr have responded to Adapa’s comment in the other thread. If I were Ewert, looking at what’s going on, I’d see a Free-Fire Zone. Would you please move the three off-topic comments here?

  6. Tom English:
    Elizabeth,

    Keiths previously created a thread for fifthmonarchyman, who wanted to respond to the question I’d posed to Ewert. I thought that Adapa’s comment was worthwhile, in and of itself, but irrelevant to my question of model reification. Now you and Piotr have responded to Adapa’s comment in the other thread. If I were Ewert, looking at what’s going on, I’d see a Free-Fire Zone. Would you please move the three off-topic comments here?

    Tom, let me know if I’ve sorted this out correctly, or if anything else needs moving.

  7. [latexpage]Elizabeth:

    “show up more often that chance would lead us to expect” is just the eleP(T|H)ant again.

    What do you mean by that? Unlike specified complexity, “active information” increases as the probabilistic complexity (exogenous information) P(T|H) increases.

    Any process with a non-flat probability distribution is going to be “biased” in favour of some outcomes more than others.

    Careful here… There’s a difference between the distribution of the process and the distribution of the outcome. Also, although I put most of the terms of Dembski, Ewert, and Marks in scare quotes, I’m not inclined to do so with bias. “Active information” is reasonably described as a measure of bias in the distribution of outcomes, relative to a baseline distribution.

    Is Active Information simply the non-flatness of the probability distribution?

    Definitely not. Let’s say that $P$ is a uniform probability distribution on finite $\Omega,$ and that $Q$ is a non-uniform distribution on $\Omega.$ It is nonetheless possible that for a subset $T$ of $\Omega,$ $Q(T) = P(T),$ i.e., the “active information” $$I_+^T(Q|P) = \log \frac{Q(T)}{P(T)} = 0.$$ I’ll mention that the log transformation is worse than useless.

  8. Elizabeth: Tom, let me know if I’ve sorted this out correctly…

    Excellent. Thank you. I won’t normally worry about stuff like this. It’s all about creating a good environment for Winston.

  9. Tom English:
    Elizabeth:

    What do you mean by that? Unlike specified complexity, “active information” increases as the probabilistic complexity (exogenous information) P(T|H) increases.

    Winston wrote:

    The target is effectively the measuring stick. The choice of target is arbitrary, and I could have as easily chosen cities, paintings, beetles, cows, volcanoes, mountains, lakes, or crystals. The same conclusion applies to all of them: they show up far more often than chance would lead us to expect. Within a materialist framework, they have to be explained by a process biased in favour of producing them.

    Which looks to me like a restatement of the p(T|H) quantity: he is saying that “within a materialist framework” the probability of any from a class of things that includes “cities, paintings, beetles, cows, volcanoes, mountains, lakes, or that have been specified specified as “Targets” out of all possible alternative configurations, given some “chance hypothesis that takes into account Darwinian and other material mechanisms” has to be not small. And that class of Targets to me looks as though it was selected by some kind of “specified complexity” criterion (extensive patterns that are simple to describe), although of course the problem with that criterion is that we also have to exclude patterns that are generated by a known material process, so the argument is circular. And in any case, we know, and Winston agrees, I think, and so does Dembski, that evolutionary processes can produce such things, hence the shift to “Search for a Search”.

    Careful here… There’s a difference between the distribution of the process and the distribution of the outcome. Also, although I put most of the terms of Dembski, Ewert, and Marks in scare quotes, I’m not inclined to do so with bias. “Active information” is reasonably described as a measure of bias in the distribution of outcomes, relative to a baseline distribution.

    Yes indeed. I’m perfectly happy with “bias”, as long as the accounting is good.

    And yes, I take your point that with Active Information, we are now discussing the distribution of processes, not the distribution of outcomes of those processes. So to rephrase: is Active information, then, the degree to which there is a bias in favor of processes in which Targets are probable outcomes?

    I’ll mention that the log transformation is worse than useless.

    Yep.

  10. Winston has been immensely helpful to a computational/bioinformatics research project I’ve been working on, and I’m greatly in debted to him, and I will say so publicly.

    That said, I have sharp difference of opinion on the utility of the information framework as it pertains to evolutionary and OOL questions.

    Thank you to Winston for critiquing things like Avida and Tierra and all his computational help to me on bioinformatics research, but I wish energy were devoted outside of information theory framework. Information theory makes sense when there is a well-defined design or project in question or when there is a defined need for a human observer (like an observer calculating how much memory storage he needs for an application). Critique of evolutionary theory, imho, is more successful at the basic level of chemistry and population genetics.

    So my personal thanks to Winston, and I wish him well. If any have wondered my silence on the technical issues of late on information theory, it is because I think there are more fruitful approaches than information theories to demonstrating the inadequacy of prevailing mainstream explanation for the architecture of biology.

Leave a Reply