login about faq

This question began with the following quote from John Paquette:

Deduction can establish true conclusions, which can then be used as premises from which further reasoning can proceed. What deduction cannot do is form new generalizations which aren't yet implicit in old generalizations.

For example: All female robins are birds. All female birds lay eggs, therefore all female robins lay eggs.

The deductive conclusion above is a generalization, but deduction only arrives at it by narrowing the wider generalization "All female birds lay eggs."

Deduction finds specific truths from general truths. Induction finds general truths from specific truths.

The following response came in:

How did we determine that all female birds lay eggs without examining female robins? If we concluded that "all female birds" have some attribute without examining female robins, how did we determine that female robins are female birds?

The Wikipedia definition is "Birds (class Aves) are feathered, winged, bipedal, endothermic (warm-blooded), egg-laying, vertebrate animals." If we go by this, then in the process of determining that a robin is a bird, we determined in some way that it lays eggs.

The question becomes: Under this formulation of deduction and induction, can deduction produce any "specific truths" that were not used via induction to find our "general truths"?

asked Apr 02 '13 at 03:12

Humbug's gravatar image


edited Apr 03 '13 at 11:35

Greg%20Perkins's gravatar image

Greg Perkins ♦♦

You can learn a lot about birds qua birds before you ever learn about the existence of a specific kind of bird called "robin". You may never have seen a robin before making the (presumably) valid induction that female birds lay eggs. Induction does not require exhaustive knowledge of the class of entities (e.g. birds) being induced from.

Deduction tells you, specifically, that your general conclusion about birds applies to the new subclass of birds, "robin", that you have just learned about.

Deduction: THIS is one of THOSE. Qua one of THOSE, X is true of it. Robins are birds, qua birds, robins lay eggs.

answered Apr 02 '13 at 16:15

John%20Paquette's gravatar image

John Paquette ♦

edited Apr 02 '13 at 16:16

"You can learn a lot about birds qua birds before you ever learn about the existence of a specific kind of [X] called robin"

In order to properly classify a robin as a bird, don't we have to use induction and validate whether it meets all the essential characteristic of a bird? Given that egg-laying is an essential characteristic of a bird, aren't we discovering that robins lay eggs via induction PRIOR to making a deduction?

In fact, the deduction doesn't come into play unless we take on (as faith?) someone else judgment that a robin is a bird and then use that as a premise for the deduction

(Apr 02 '13 at 17:38) Humbug Humbug's gravatar image

This question pertains to the role of deduction in gaining knowledge. The topic is clouded somewhat by terminology such as "produce truth" (truth is discovered and identified, not "produced"), and by a distinction between "specific truth" and "general truth" that isn't clearly defined unless "general" just means a class, and "specific" means a class member or instance.

As the question points out, the egg-laying example has a basic problem: egg-laying is arguably an essential defining characteristic of birds, in which case one needs to examine Robins first, before concluding that they are birds, to determine whether or not they reproduce by laying eggs. On this view, if female Robins don't lay eggs, then they're not birds.

But suppose the example is changed a little, as follows:

  • Suppose modern science has established that any organism that has attributes 'x', 'y', and 'z' also has attribute 'w'. Suppose science has established this not only by studying a huge number and wide variety of actual cases, but also by identifying a causal connection between the attributes such that 'w' causes and necessitates 'x', 'y' and 'z' (or vice versa).

  • Supose a new type of organism is found that has attributes 'x', 'y', and 'z', but we don't know in advance if it also has 'w'. To determine whether or not it has 'w' by examining it directly, assume one would need highly expensive test equipment and very time-consuming testing, financially prohibitive without a bigger budget. Can we nevertheless conclude that this new organism has attribute 'w', just based on having found that it has 'x', 'y' and 'z'? Surely we can say that it very probably (almost certainly) has 'w'. Does that mean we know that it has 'w'? A university research professor might add the organism to a prioritized list of topics for future study when time and resources allow, and a sufficiently interested graduate student might become intrigued by the thought of becoming the first ever to find an organism that has 'x', 'y' and 'z' but not 'w'. Until the further research is actually done, however, we can only say that 'w' is "highly probable" (even "virtually certain") in the new organism, but not yet confirmed by direct observation.

This example illustrates Objectivism's basic perspective on how man gains knowledge:

  • "Knowledge" is knowledge of reality, i.e., of existents and existence. It is knowledge of that which exists. So-called "knowledge" of anything that is not real is only imagination (at best), formed ultimately by a process of mentally rearranging elements that do exist or could exist.

  • Man gains knowledge of reality primarily by looking at reality, identifying what he finds and integrating it with all his other knowledge of reality. It is specifically in the "integration" phase that deductive logic is most applicable and valuable. Logic, defined as "the art [or cognitive process] of non-contradictory identification," is enormously helpful to man in discovering errors in his integrations leading to contradictions, since contradictions cannot exist (because A is A). Contradictions prevent integration; resolving them allows integration to proceed. In this way, logic greatly helps to validate and sharpen one's integrations.

  • All knowledge is contextual. It arises within a definite context of facts and prior knowledge, and, once validated, is known with certainty in that context. But if the context expands through new discoveries, the knowledge that was valid in the former context (and still is valid in that context) may need to be amended and expanded to accurately describe the expanded context.

OPAR explains the role of logic in Chapter 4, "Objectivity," in the section titled, "Objectivity as Volitional Adherence to Reality by the Method of Logic." Logic is not a means of "producing new truths." It is a tool to assist man in the process of integrating his observations and immediate identifications. (Refer also to the sections titled, "Concepts as Objective," "Knowledge as Contextual," and "Knowledge as Hierarchical." For contrast, the chapter ends with "Intrinsicism and Subjectivism as the Two Forms of Rejecting Objectivity.")

The ways in which specific observations and identifications may be connected to other observations and integrations are not necessarily self-evident without cognitive effort. The process of integration -- and of seeking out and correcting any contradictions -- can often reveal connections that one may not initially have comprehended, thereby broadening and strengthening one's integrations. Logic doesn't "produce" new knowledge; it only helps man to see connections and possibilities (or clashes) he may not have recognized previously.

Regarding birds that may or may not lay eggs (i.e., types that produce offspring by laying eggs), it is possible for a new concrete to fit within a previously formed concept in terms of the essential distinguishing characteristics of the units of the concept, yet differ from the other units in some significant way. An type of animal that looks like a bird and acts like a bird but doesn't lay eggs like a bird would be an example. Such an exception case might well turn out to be sufficiently different from birds to warrant a new classification of its own, for a type of animal not previously known. The mere fact that it matches an existing concept in all ways previously known and identified as fundamental to the concept does not prove or mandate that the new type of concrete must match in other respects. Whether or not it does match, and how closely, is for man to observe and identify by looking at it directly and more closely. Reality always is what it is, and man's concepts need to be adjusted and adapted accordingly as the evidence warrants.

(Technically, egg-laying cannot be an essential definining characteristic of "bird" because half of the bird population doesn't qualify, namely, the male half. And even among females it is not necessarily true that all females actually perform egg-laying. Some may fail to find a mate and build a nest. What the females possess, if they are not defective, is an egg-laying capacity, which may or may not be fully functional in any particular female.)

Update: Certainty

A comment asks:

Are you saying that you can be certain of a particular piece of knowledge even though you can't be certain that your knowledge won't change if your context changes in the future?

The answer is yes, exactly. It should also be added that the kind of context change involved here normally happens very infrequently and/or slowly. It's not something one normally needs to make a special effort to remember and acknowledge. One need not go around continually qualifying one's claims as subject to a specific context if the context and the dependency are already understood.

OPAR explains the Objectivist view of certainty and absolutes in Chapter 5, "Reason," in the section titled, "Certainty as Contextual." Pages 172-174 observe:

Man is a being of limited knowledge -- and he must, therefore, identify the cognitive context of his conclusions. [...]

If a man follows this policy, he will find that his knowledge at one stage is not contradicted by later discoveries. He will find that the discoveries expand his understanding; that he learns more about the conditions on which his conclusions depend; that he moves from relatively generalized, primitive observations to increasingly detailed, sophisticated formulations. He will also find that the process is free of epistemological trauma. The advanced conclusions augment and enhance his earlier knowledge; they do not clash with or annul it. [...]

The appearance of a contradiction between new knowledge and old derives from a single source: context-dropping. If the researchers [who discovered blood types] had decided to view their initial discovery as an out-of-context absolute; ... in effect, as a matter of dogma... then of course the next [blood] factor discovered would plunge them into contradiction, and they would end up complaining that knowledge is impossible. [...]

Although the researchers cannot claim their discovery as an out-of-context absolute, they must treat it as a contextual absolute (i.e., as an immutable truth within the specified context). [...]

The fact of context does not weaken human conclusions or make them vulnerable to overthrow. On the contrary, context is precisely what makes a (properly specified) conclusion invulnerable.

This OPAR discussion goes on to explain further what is meant by "absolute" (p. 175):

The modern definition of "absolute" represents the rejection of a rational metaphysics and epistemology. It is the inversion of a crucial truth: relationships are not the enemy of absolutism; they are what make it possible.

Further overview of the Objectivist perspective on certainty and absolutes can be found in The Ayn Rand Lexicon under the topics of "Certainty" and "Absolutes." There is also illluminating discussion by Ayn Rand in ITOE, 2nd Ed., Appendix, section titled, "Scientific Methodology."

answered Apr 05 '13 at 00:13

Ideas%20for%20Life's gravatar image

Ideas for Life ♦

edited Apr 10 '13 at 22:12

What is "almost certain" and "virtually certain"? Are you saying that you can be certain of a particular piece of knowledge even though you can't be certain that your knowledge won't change if your context changes in the future?

(Apr 08 '13 at 03:00) Humbug Humbug's gravatar image

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here



Answers and Comments

Share This Page:



Asked: Apr 02 '13 at 03:12

Seen: 1,267 times

Last updated: Apr 10 '13 at 22:12