Deduction and Induction Flashcards

1
Q

Deduction - Aristotle

A
  • Regarded as one of the most long lasting and influential contributions made to civilisation.
  • Other systems include Stoicism
  • Kant thought that Aristotle had discovered all there was to know about logic.
  • The term ‘Traditional Logic’ is used to describe the offspring of Aristotelian logic.
  • The ‘book’ that contains Aristotle’s extant writings on logic is known as the Organon. This is in fact a set of 5 or 6 writings on the topic.
  • Aristotle while acknowledging Induction has much less to say about it than Deduction.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Syllogistic reasoning or logic

A
  • Comes from greek for deduction
  • Aristotle outlined the basic structure consisting of accepting things that are certain from the known facts.
  • The structure of deductive statement is thus in two parts
    • The premises (2) and
    • The conclusion
  • Syllogisms are thus heavily reliant on two elements;
    • The ‘known-ness’ of the facts
    • The linguistic understandings of the reasoners
  • Two types of syllogisms in Aristotle’s system:
    • Linear
    • Categorical
  • Linear less important for our current purpose but have a form of a linear argument. These are somewhat superfluous and add little to what is immediately obvious.
  • E.g.
    • You are taller than John;
    • John is taller than Bill;
    • Thus you are taller than Bill
  • Categorical syllogistic reasoning
    • Consists of two premises and a conclusion that automatically follows
    • There are four types which form a 2x2 matrix. Based on two aspects: membership and coverage
    • The terms for membership are “affirmation” and “negation”
    • For coverage “universal” and “particular”
    • Syllogisms have specific words which indicate the type of argument being put forward;
      • Affirmative universal is “ALL”
      • Affirmative particular “SOME”
      • Negative universal “NO”
      • Negative particular “SOME…NOT”
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Conditional reasoning

A
  • A conditional reasoning proposition consists of an antecedent and a consequent
  • They are of the form If p, then q where p is the antecedent and q is the consequent
  • What follows from the proposition is a re-statement of the implications of the conditional proposition
  • Two prominent
  • Modus ponens or affirmation of the antecedent
    • If p then q
    • Given q
    • Deduce p
  • Modus tollens
    • If p then q
    • Not q
    • Not p
  • There are a number of other statement types which have been studied and found to be problematic from the pov of human performance.
  • One of the most famous is the Wason selection test.
    • Takes the form of a set of cards with an associated proposition, rule or question. E K 4 7.
    • The responses of participants are invariably poor by standards of deductive logic on average between 4-10%
    • To test the truth of the claim or rule logic requires that participants examine the assertion or p and the negation of the consequent i.e not q
    • The performance is seen to improve dramatically when relevant content or context is introduced. For example the drinking rule.
    • This raises the issue of to what extent is human reasoning context dependent and so abstract tasks are of minimal use. The latter form is called deontic reasoning.
  • Theories put forward to account for deductive reasoning
    • Rips Mental Rules
      • Asserts that people solve deductive problems by constructing mental proofs
      • These proofs consist of a set of sentences in working memory
      • These sentences link premises to conclusions by inference rules
      • Not following the rules occurs as a result of cognitive limitations such as working memory capacity.
    • Johnson Laird’s Mental Models
      • People construct mental models based on the premises of the arguments
      • These models relate to real life knowledge
      • People then attempt to solve the problems by attempting to construct alternatives. If none can be created then the initial model is accepted
      • Mental models research is generally considered the more comprehensive and robust.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Why study reason deductively?

A
  • Evans asserted in 1982 that deduction was capable of being stripped of its association which were generated by experience, training and human adaptability
  • Deductive problems in a sense allow as pure a form of experimentation s possible
  • This however again raises the question of normative or descriptive focus. It is appropriate to study human reasoning by placing it outside of its ecological niche?
  • The findings of Deontic experiments make the point that human reasoning can follow the rules of Deductive logic to a great extent provided the context is included.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Summary - deductive reasoning

A
  • There are many forms of deductive reasoning. The most commonly studied are
    • Syllogistic reasoning
    • Conditional reasoning
  • One of the great strengths of deductive reasoning is that if the arguments are said to be true then the conclusion is guaranteed
  • Deduction does not take us beyond the assertions or premises. This is the great weakness, one cannot learn anything new.
  • Syllogistic reasoning involves premises followed by a conclusion
  • Conditional reasoning involves if-then statements with the form of antecedent followed by consequent
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Induction - general

A
  • Induction is said to be “an activity of the mind that takes us from the observed to the unobserved”
  • Induction involves “a continuous adjustment to and updating of ones confidence in a belief”
  • Inductive reasoning is the most common form of everyday and scientific reasoning
  • Induction is the method of human thinking can generate assertions beyond the data or premises that are used in the process of inductive reasoning.
  • It is by induction that we move from ‘sample’ to ‘population’
  • Induction is associated with hypothesis generation and testing
  • Induction involves the continuous adjustment of one’s beliefs
  • One of the pitfalls is that in going from the particular to the general a cast number of hypotheses can be generated.
  • There are two types of inductive assertion:
    • Strong: this is where there is good reason to accept the assertion under consideration
    • Weak: here there is poor or weak support [if there is no support than we should not accept the assertion]
  • Induction is one of the main topics that Hume’s critique of causation applies to.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Induction - causal inference

A
  • One of the main assertions made by Hume was that the concept of causality was the product of co-occurrence in the experience and therefore mind of the observer i.e. us
  • Causation is “mental habit formed by past regularities”
  • Induction is the main method of deriving causal relations. We observe “many” instances and so the events become paired. A causes B. Flame causes heat.
  • Such a method is said to be a critical capability in pre-scientific method cultures and hence for early humans.
  • By pairing regularly occurring events we gradually develop a model of the world and what is safe and not and what is neutral.
  • Our degree of association is a result of the likelihood of events co-occurring. In other words the probability of events co-occurring is critical to rational belief.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Induction - confirmation bias

A
  • The use of induction as a method for assessing the implications of events in the world take an interesting turn. Humans are prone to belief persistence or resist belief change.
  • This has shown itself in several experimental findings around the topic of confirmation bias.
  • Example
    • Numbers challenge task from 1st workshop.
  • Many results support these findings. People are much more likely and it seems to come “naturally” to ask confirmatory questions and not disconfirmatory ones.
  • These studies have been criticised on several grounds including influence of social norms, source credibility and participants may have been affirming the hypothesis.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Abduction

A
  • Concept used in philosophy to indicate the process of hypothesis generations
  • More exactly it means that selection of the best explanatory hypothesis for a known set of facts
  • Is almost the reverse of induction in that the conclusion is accepted and the process involves the selection of the best explanatory set of premises.
  • It is applied in such areas as fault diagnosis in automated systems, belief revision which is a hallmark of scientific reasoning.
  • Abduction may be seen as incorporating new information at the level of ordering of the preference of the possible worlds.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly