Home » Intelligent Design » The Design Inference now in paperback

The Design Inference now in paperback

My book The Design Inference is now out in paperback (I just received 6 copies via FedEx from Cambridge University Press). It might interest readers of this blog to see the difference in the back covers between the paperback edition and the original hardcover edition (after the first two printings, Cambridge omitted the jacket cover of the hardback edition, so it is no longer widely available):

Paperback edition cover in back:
TDI paperback backcover

Hardback edition cover in back:
TDI hardback backcover

  • Delicious
  • Facebook
  • Reddit
  • StumbleUpon
  • Twitter
  • RSS Feed

22 Responses to The Design Inference now in paperback

  1. Bill,

    Perhaps you could apply the explanatory filter to the typo on the back to determine whether it’s there by design or accident!

    I would have thought Cambridge University Press would have a better proofreader.

  2. It looks like they threw in the bullets just to fill out the cover, mindless of the repetition immediately above.

    I take it they didn’t submit the covers to you for approval?

  3. William, congratulations on your new book!

    Typo is nothing “Provides a solution [t]o the long standing problem,” but the other thought is hard to understand – “Just about anything that happens is highly improbable.”

    I asked you once before in this blog, in view of frequency interpretation of probability “highly improbable events happen by chance all the time” means “Highly infrequent events happen frequently.”

    Can you explain your claim?

    previous post: http://www.uncommondescent.com.....1#comments

  4. doctormark writes:
    “What a shameless self-promoter. Whatever happened to “Christian” modesty?”

    doctormark,
    I don’t think he was boasting. I think he was lamenting the way they mangled the cover with the glaring typo and the oafish repetition of the second paragraph by the bullet items.

  5. I know! That’s totally unchristian like to share knowledge with people…paperback versions are cheaper and more readily available to more people. Christ would have surely banished Bill from the kingdom for such behavior, no doubt!

  6. BTW. Why did they remove the two quotes, opting for the bullet points in their place? Is there a reason behind that?

  7. es writes:
    “I asked you once before in this blog, in view of frequency interpretation of probability ‘highly improbable events happen by chance all the time’ means ‘Highly infrequent events happen frequently.’”

    es,
    Look at a state lottery. Each particular set of winning lottery numbers is highly improbable, yet a set of numbers is drawn every day. So in this case, a (different) improbable event happens every day. In other words, “Highly improbable events happen by chance all the time.”

    To put this correctly into frequency terms, you’d have to say “It’s frequently true that an actual event comes from a set of many events, each of which is highly infrequent.”

  8. Keith,

    > “It’s frequently true that an actual event comes
    > from a set of many events, each of which is
    > highly infrequent.”

    There is nothing improbable about such an event. That is why we observe such events frequently.

  9. es,

    Let me try again.

    It frequently happens that an improbable event occurs.
    It infrequently happens that a particular improbable event occurs.

    See the difference?

    By the way, there’s nothing about this claim that is specific to intelligent design, in case you’re wondering.

  10. keith,

    Let me try again as well. The sentence is a contradiction – “It frequently happens that an improbable event occurs.”

    Improbable events do not occur frequently, probable events do. This is what frequency interpretation is all about.

    Let’s take a very simple example, coin tosses. You toss a coin 500 times, you obtain some sequence. Probability of the head coming up is 0.5, same for the tail.

    The probability of such event is:

    (0.5+0.5) x (0.5+0.5) x … [500 hundred times total] = 1

    You can easily verify (even in a mental experiment) that every time you toss a coin 500 times you will obtain some sequence, with certainty, thus a probability of 1.

    But if someone will give you a piece of paper with the sequence to be obtained before such an experiment you can be absolutely sure (within our current knowledge of physical laws, timespan of the universe and so on) that there is some form of cheating.

    Highly improbable events cannot happen all the time. What can happen all the time is that we by mistake assign low probability to a probable event.

  11. I was watching the Annenberg CPB channel yesterday and they were discussing statistics and the art of it in various fields…anyhow- this guy had practiced flipping a coin a certain speed and angle where he could get it to show heads everytime.

    I had to wonder why he had so much free time. Probably doesn’t date a lot. :)

  12. es writes:
    “The sentence is a contradiction – ‘It frequently happens that an improbable event occurs.’”

    es,
    I think the misunderstanding here has more to do with English usage than with probability. Are you, by any chance, a non-native speaker of English?

    Consider the following two English sentences:
    1. A pedestrian is unlikely to be killed by a car when crossing the street.
    2. A hundred times a day, a pedestrian is killed by a car when crossing the street.

    Both are perfectly acceptable sentences in English, with sensible meanings, and they are not the least bit contradictory. But a consistent interpretation of the phrase ‘a pedestrian’ would lead you to believe these two sentences are in fact contradictory. Of course, in sentence #1 the phrase ‘a pedestrian’ means ‘a particular pedestrian’, while in sentence #2 it refers to ‘one of the set of all pedestrians’.

    Analogously to sentence #2, the sentence ‘It frequently happens that an improbable event occurs’ is understood to mean ‘It frequently happens that one of a set of improbable events occurs.’

    Is it clear yet?

  13. Keith, you are correct, English is not my native tongue.

    I wonder, if the sentence is translated to some other language, is your argument still valid? Or do you think the meaning of the sentence can only be understood in English?

    I thought we were discussing probability math, not idioms. Do you have any arguments against what I posted in #11?

    P.S. By the way, are you a native speaker? This explanation – “Of course, in sentence #1 the phrase ‘a pedestrian’ means ‘a particular pedestrian’” is not clear to me. Why not use ‘the pedestrian’ if you are referring to someone in particular?

  14. Wow, this dialog is interesting at least. But why I really wanted to comment was to say congrats on the book release. A highly improbable event that happens quite frequently. (See, most people never get published – even highly improbable that someone will get published, and yet there are thousands of books published each year!) :-) lgp

  15. es wrote: “By the way, are you a native speaker? This explanation – “Of course, in sentence #1 the phrase ‘a pedestrian’ means ‘a particular pedestrian’” is not clear to me. Why not use ‘the pedestrian’ if you are referring to someone in particular?”

    I am a native speaker of English, but am envious of anyone who speaks more than one language.

    If I may presume to explain, Keith’s example of “a pedestrian” means more accurately “any given pedestrian” or “any particular pedestrian” or yet again, “any individual pedestrian”.

    In Keith’s example “the pedestrian” would be a bit confusing because it is TOO specific. It gives the impression that the speaker and hearer know WHO the pedestrian is or they are referring to some previously referenced pedestrian. But we don’t know who the pedestrian is, nor do we really care. We just know that it is one particular pedestrian whose identity is unimportant.

    Whew! It’s much harder trying to explain the WHY of a language than to actually use it correctly.

    This is probably confusing, but I gave it a shot.

  16. russ,

    Thanks for giving it a shot. Have you ever thought so hard in your life about what the word “a” means? I know I haven’t.

    es,

    I was working on a reply when russ posted his. I agree with him, but I’ll go ahead and post mine because it contains several examples that might help you understand what we’re both trying to tell you:

    Yes, I am a native English speaker, though the British might claim I speak “American”, not “English.” :-)

    You ask a good question, and it’s not one that I normally have to think about. As a native speaker, I automatically know what “sounds right” and what “sounds wrong.”

    Let me try to explain it, though. Consider these four sentences:

    1. The pedestrian was hit at the corner.
    2. A pedestrian was hit at the corner.
    3. Twenty times last year, a pedestrian was hit at the corner.
    4. Twenty times last year, the pedestrian was hit at the corner.

    In sentence number #1 we are speaking of a particular pedestrian who was specified in a previous sentence. In sentence #2 we just mean a single pedestrian who has not previously been specified.

    To see how this works, read the following paragraph, and watch for how ‘a pedestrian’, ‘a driver’, ‘an accident’ and ‘a police officer’ change to ‘the pedestrian’, ‘the driver’, ‘the accident’ and ‘the police officer’ once we have a particular one in mind:

    Last night there was an accident. A pedestrian was walking down Main Street. A driver was speeding down 1st street toward Main street. The pedestrian began to cross the street. The driver failed to stop. The pedestrian was hit at the corner. I reported the accident to a police officer. The police officer called an ambulance.

    Sentence #3 means that there were twenty pedestrian/vehicle collisions at the corner last year. The pedestrian and vehicle were not the same each time, but each time there was one pedestrian and one vehicle. Sentence #4 means that some poor soul (let’s call him ‘Richard D.’ to satisfy the pro-ID folks on this blog) was hit at that corner twenty times in the last year. The victim was Richard D. in all twenty cases.

    Your math in comment #10 is correct. The way I would express it in English is:
    5. The probability of getting a sequence is 1. (“a sequence” = “any sequence at all”)
    6. The probability of getting a particular sequence is tiny.

    Back to the original sentence that confused you:
    “Just about anything that happens is highly improbable.”
    This can be restated as:
    Almost every event that occurs comes from a large set containing many improbable events. Each individual event has a low probability, but the probability of getting an event from the set as a whole is high.

    I hope that helps.

  17. Russ,

    Thanks for your explanation, I understand what you are saying and agree that hypothetically on average it will be unlikely for any arbitrary pedestrian to be killed when crossing the street. (Divide the number of deaths in a day by the number of pedestrians, right?)

    Accepting this as a fact, I don’t see any contradiction with another hypothetical fact that a hundred times a day a pedestrian is killed – it all depends on number of pedestrians and other things.

    But the analogy that Keith is asserting

    > Analogously to sentence #2, the sentence
    > ‘It frequently happens that an improbable event
    > occurs’ is understood to mean ‘It frequently
    > happens that one of a set of improbable events occurs.’

    I believe is incorrect. Words certainly can have many meanings (“The question is,’ said Alice, `whether you can make words mean so many different things.’ `The question is,’ said Humpty Dumpty, `which is to be master — that’s all.’”)

    There are different probabilities for A. any event out of a set events to happen and B. any specific event from the set to occur.

    The probability of obtaining any sequence of 500 coin tosses is 1. The probability of obtaining any specific (specified) sequence is 2^-500. We can fine tune, for example, specify that the first and the last tosses should be heads, the rest may have any value (probability then is 0.25) and so on, but we cannot use the same designation/description for vastly different probabilities, for it leads to confusion and false reasoning.

  18. Keith, thanks, I think we cleared the indefinite article usage, I take your word for it. (I thought it was funny, considering ‘particularizing effect ‘ in the definition of ‘the’:

    > 1. (used, esp. before a noun, with a specifying or particularizing effect, as opposed to the indefinite or generalizing force of the indefinite article a or an): the book you gave me; Come into the house.
    http://www.infoplease.com/ipd/A0691136.html

    But back to our topic,

    > Almost every event that occurs comes from a large set
    > containing many improbable events. Each individual
    > event has a low probability, but the probability of
    > getting an event from the set as a whole is high.

    The distinction that I argue for is that we cannot use the probability of getting *any* event from the set in place of the probability of getting a *specified* event.

    If we can agree on this, then I would like to hear an explanation for “exceedingly improbable things happen all the time.” How do we know they are exceedingly improbable?

    Cheers

  19. es,

    Okay, one final attempt to explain this. If this doesn’t work, I think you’ll need to find someone who speaks both your native language and English fluently, and who understands probability. By the way, what IS your native language?

    “…I would like to hear an explanation for ‘exceedingly improbable things happen all the time.’”

    Imagine you spill a tablespoon of salt on a table. The exact arrangement of salt grains is exceedingly improbable. You could spill salt over and over again for the rest of your life and never get the same arrangement. Every time you spill the salt you get an enormously improbable arrangement. Many events. Each one highly improbable. Yet people spill salt thousands of times a day across the world. It happens all the time. Therefore, exceedingly improbable things happen all the time.

    How do I know a particular arrangement is improbable? It never happens again, for as long as I have the patience to keep spilling the salt. Low frequency equals low probability.

    It’s extremely unlikely that you or I or any particular person will be hit by lightning. But people get hit by lightning all the time. Each one of them was extremely unlucky to be hit by lightning. Different events. All improbable. Exceedingly improbable things happen all the time.

    That’s the best I can do, es. If you tell us your native language, perhaps someone who is lurking on the blog can help you out.

  20. Sorry to post somewhat off-topic. I’m a new user, and my earlier comments were delayed in moderation I think and got a bit lost on earlier threads. I’m a physicist, and a bit skeptical about ID. I’ve basically got five questions which summarise my lack of understanding, and I’d be interested in any comments. I understand that much of this may be discussed in the book above, so sorry if this is old ground.

    As I understand it, the two possible mechanisms for evolution discussed on
    these pages (and elsewhere in the ID debate) are:
    (1) unguided (Darwinian) evolution and
    (2) guided (ID) evolution.

    Both mechanisms will result in apparent design to some extent, (2) by
    definition and (1) by natural selection. The important question is
    experimental, i.e. whether one or both or neither mechanism is consistent
    with the designs we observe.

    The experimental data is then the design and functionality of biological
    systems. My first question is, are there possible designs one could find
    in nature which could not be explained by mechanism (2)? I think the
    answer is assumed by many scientists to be `no’, which is why (2) is often
    regarded as non-falsifiable.

    On the other hand, much of the evidence in the ID debate is I think
    intended as a putative falsification of (1). I’m not an expert in deciding
    whether a particular design found in nature is consistent with mechanism
    (1), but as I understand it the main argument is that some designs are
    irreducibly complex. My sense is that rigorously proving that there is no
    pathway to produce a particular design via random mutations is a hard
    proof. Can IC be proved rigorously? Most biologists seem not agree with
    the proposed cases.

    My third question is: does the falsification of (1) imply the veracity of
    (2)? I.e. can the ID hypothesis be defined as the complement of a previous
    theory, (1)? This unusual in science—usually there is some possibility
    that two competing theories are both inconsistent with the data. I think
    this is another reason (or perhaps the same reason, rephrased) that
    ID is sometimes seen as non-falsifiable.

    My final point/question is that, as a physicist, I see biological
    processes (in a reductionist sense) as very low-energy physics. As physicists we feel we understand very well how these low energy physics
    processes work, in principle. Does (2) entail physical laws being changed? Are
    current mutations of lifeforms thought to be guided? Could we look
    for these changes?

    Best wishes, and I’d be really interested in your responses. To summarise with the five questions:

    (a) Are there possible types of design which would be inconsistent with
    guided evolution?

    (b) Is the evidence for irreducible complexity in specific cases convincing enough to
    falsify unguided evolution? How does one prove IC rigorously in general?

    (c) Is falsification of (1) equivalent to proof of (2)?

    (d) If (c) is true, and I think by implication (a) is also true, then is mechanism (2) a sensible scientific theory? I think it is unprecedented that one could `prove’ one
    scientific theory by the very process of falsifying another.

    (e) Is guided evolution ongoing, and does this mean we should observe
    different low-energy physics?

  21. My first attempt at posting this didn’t go through. Trying again…

    es,

    Okay, one final attempt to explain this. If this doesn’t work, I think you’ll need to find someone who speaks both your native language and English fluently, and who understands probability. By the way, what IS your native language?

    “…I would like to hear an explanation for ‘exceedingly improbable things happen all the time.’”

    Imagine you spill a tablespoon of salt on a table. The exact arrangement of salt grains is exceedingly improbable. You could spill salt over and over again for the rest of your life and never get the same arrangement. Every time you spill the salt you get an enormously improbable arrangement. Many events. Each one highly improbable. Yet people spill salt thousands of times a day across the world. It happens all the time. Therefore, exceedingly improbable things happen all the time. This doesn’t mean that a particular improbable thing happens all the time, as you seem to be objecting.

    How do I know a particular arrangement is improbable? It never happens again, for as long as I have the patience to keep spilling the salt. Low frequency equals low probability.

    It’s extremely unlikely that you or I or any particular person will be hit by lightning. But people get hit by lightning all the time. Each one of them was extremely unlucky to be hit by lightning. Different events. All improbable. Exceedingly improbable things happen all the time.

    That’s the best I can do, es. If you tell us your native language, perhaps someone who is lurking on the blog can help you out.

  22. Keith, changing examples does not accomplish anything. A sequence of any 500 coin tosses is not repeatable either, and you agreed before: “Your math in comment #10 is correct.” Your reasoning is wrong in the latest post.

    Thanks for your help, but please spare me your condescension. I was not really asking you, come to think of it.