7873
581bf290-df83-4265-8fe4-b47b91208407
The Design Inference: Eliminating Chance through Small Probabilities
Dembski, William A.
calibre (7.6.0) [https://calibre-ebook.com]
1998-09-13T00:00:00+00:00
<div>
<p>cited in <a href="https://philos-sophia.org/about-the-film/"><em style="color: #2980b9">The End of Quantum Reality</em></a> biography-documentary of Dr. Wolfgang Smith and his <a href="https://isidore.co/calibre#panel=book_details&book_id=7678"><em style="color: #2980b9">Vertical Causation</em></a></p>
<hr>
<p>The design inference uncovers intelligent causes by isolating their key trademark: specified events of small probability. Just about anything that happens is highly improbable, but when a highly improbable event is also specified (i.e. conforms to an independently given pattern) undirected natural causes lose their explanatory power. Design inferences can be found in a range of scientific pursuits from forensic science to research into the origins of life to the search for extraterrestrial intelligence. This challenging and provocative 1998 book shows how incomplete undirected causes are for science and breathes new life into classical design arguments. It will be read with particular interest by philosophers of science and religion, other philosophers concerned with epistemology and logic, probability and complexity theorists, and statisticians. **</p>
<hr>
<p><a href="https://mathscinet.ams.org/mathscinet-getitem?mr=1650908"><span style="color: #2980b9">MR1650908</span></a></p>
<p>The most famous "design argument'' is the argument for the existence of God from the lawfulness of the universe and the intricateness of nature, but the type of argument has wider application, as when we see marks on the beach that form the pattern CONSTANTINOPLE and infer that some human—possibly someone with an acquaintance with the history of statistics—has visited the beach since the last high tide. <br> This book sets out to uncover the logic of the design argument. The central idea is a variation on the logic of significance testing: to reject the hypothesis that the evidence <a name="MathJax-Element-1-Frame"></a><em style="font-family: 'MathJax_Math'">E</em> is the result of any chance process (and therefore must be produced "by design'') the evidence must have very low probability under any chance hypothesis <a name="MathJax-Element-2-Frame"></a><em style="font-family: 'MathJax_Math'">H</em> and satisfy a number of further conditions: (1) conditional independence <a name="MathJax-Element-3-Frame"></a><em style="font-family: 'MathJax_Math'">P</em><a name="MathJax-Span-10"></a><span style="font-family: 'MathJax_Main'">(</span><a name="MathJax-Span-11"></a><em style="font-family: 'MathJax_Math'">E</em><a name="MathJax-Span-12"></a><span style="font-family: 'MathJax_Main'">|</span><a name="MathJax-Span-15"></a><em style="font-family: 'MathJax_Math'">H</em><a name="MathJax-Span-16"></a><span style="font-family: 'MathJax_Main'">&</span><a name="MathJax-Span-17"></a><em style="font-family: 'MathJax_Math'">I</em><a name="MathJax-Span-18"></a><span style="font-family: 'MathJax_Main'">)</span><a name="MathJax-Span-19"></a><span style="font-family: 'MathJax_Main'">=</span><a name="MathJax-Span-20"></a><em style="font-family: 'MathJax_Math'">P</em><a name="MathJax-Span-21"></a><span style="font-family: 'MathJax_Main'">(</span><a name="MathJax-Span-22"></a><em style="font-family: 'MathJax_Math'">E</em><a name="MathJax-Span-23"></a><span style="font-family: 'MathJax_Main'">|</span><a name="MathJax-Span-26"></a><em style="font-family: 'MathJax_Math'">H</em><a name="MathJax-Span-27"></a><span style="font-family: 'MathJax_Main'">)</span>, where <a name="MathJax-Element-4-Frame"></a><em style="font-family: 'MathJax_Math'">I</em> is the reasoner's background knowledge; (2) tractability of the evidence <a name="MathJax-Element-5-Frame"></a><em style="font-family: 'MathJax_Math'">E</em>. This is an idea from computational complexity: to reject the chance hypothesis you must be able to construe a description of the evidence of the basis of your background knowledge <a name="MathJax-Element-6-Frame"></a><em style="font-family: 'MathJax_Math'">I</em>; (3) delimitation: <a name="MathJax-Element-7-Frame"></a><em style="font-family: 'MathJax_Math'">E</em> should entail the description <a name="MathJax-Element-8-Frame"></a><em style="font-family: 'MathJax_Math'">D</em><a name="MathJax-Span-44"></a><span style="font-family: 'MathJax_Main'">∗</span>; and (4) <a name="MathJax-Element-9-Frame"></a><em style="font-family: 'MathJax_Math'">D</em><a name="MathJax-Span-49"></a><span style="font-family: 'MathJax_Main'">∗</span> should have low probability under the chance hypothesis. <br> The author is a fellow at the Discovery Institute's Center for the Renewal of Science and Culture in Seattle, a Christian think-tank. Except for a possible hint in a reference in the dedication to Proverbs 1, 8–9 ("Don't give up the teaching of your parents''), the book does not take a position on the soundness of the argument from design. <br> {Reviewer's remark: Laplace, who discussed the CONSTANTINOPLE example and used the term "vraisemblable'', can be interpreted as supporting a likelihood interpretation of the design argument, in the sense that he compares hypotheses by their likelihood, the probability they assign to the observation. The book does not attempt to clarify the logic of the design argument in the framework of the likelihood paradigm, although the author introduces the term "likelihood'' and even talks of a "likelihood principle''. His use of these expressions is idiosyncratic since he defines likelihood as "denoting how likely <a name="MathJax-Element-10-Frame"></a><em style="font-family: 'MathJax_Math'">E</em> is to occur under the assumption that <a name="MathJax-Element-11-Frame"></a><em style="font-family: 'MathJax_Math'">H</em> obtains and upon the condition that <a name="MathJax-Element-12-Frame"></a><em style="font-family: 'MathJax_Math'">H</em> is as effectively utilized as possible.'' Probability is seen as an estimate of "likelihood''. His "likelihood principles'' similarly are more akin to what is called "direct inference'' in the literature.} Reviewed by <a href="https://mathscinet.ams.org/mathscinet/search/author.html?mrauthid=169330"><span style="color: #2980b9">Zeno G. Swijtink</span></a></p></div>
Cambridge University Press
QA279.D455 1998
9780521623872
Zanic8M0PjgC
B00INYG5NG
776949670
eng
Cambridge Studies in Probability
Induction and Decision Theory
Experimental design
Probabilities