Lecture 4 Flashcards
Suppose we knew all events, science would still be incomplete, because we would still lack:
- awareness of connections between events
- Knowledge of events which are accidental and which are determined (some events are deeply embedded in the world)
- Ability to intervene effectively in the world
What is an explanation?
- An account that gives us understanding of why things are as they are
- Answer to why question
- Consists of explanandum and explanans
Explain the concept of explanandum and explanans
- Explanandum: that which is to be explained (answer)
- Explanans: that which does the explaining (laws)
What is requirement of an explanation?
- Truth; claims used in explanations must be true
- Otherwise, we have a pseudo-explanation (intended explanation, but it is not true)
What are the two influential models of German philosopher Carl G. Hempel?
- Deductive nomological model (DN model)
- Inductive-statistical model (IS model)
What is the Deductive-nomological (DN) model?
- An explanation is a valid deductive argument
- from true premises
- that includes at least one law (nomological) or true generalisation and descriptions of some particular facts
- to the description of the fact that is to be explained
From general rule to a specific case
–> We use the DN model to make a prediction
What is the difference between explanation and prediction?
- Premises of an explanation must be true
- but we can have a successful prediction from false premises
- We tend to use the same formula for both prediction and explanation
What is the equation of the DN model?
- law
- particular fact
- explanandum
What are two objections to the DN model?
- The DN model allows for the explanations of a cause on the basis of its effect (flagpole example)
- The DN model allows for an event to be explained based on irrelevant information
What is the Inductive-statistical (IS) model?
- Extension of DN model to probabilistic phenomena (Hempel)
- An explanation is an argument that establishes that the explanandum had high probability
- inductive, because it uses probable, not certain, reasoning
- Usually involves statistics/percentages
What are some problems with the IS model?
- Some good explanations does not make the explanandum highly likely
- The IS model is too restrictive
What was the problem with Hempel’s models?
- He thought that laws were enough to capture causality
- Placed laws at heart of the DN model
- But counterexamples of DN model show that a factor can be part of the sufficient outcome, but not the cause of that outcome
What are the conclusions of the critique on Hempels models?
- Explanations should track causes, not merely state sufficient conditions for occurrence
- DN model fails this requirement
- We need new, more causal models of explanation
What is the causal-mechanical model?
- An explanation of an event is a description of part of the causal processes and interactions that led up to said event
- Two key concepts:
1. Causal process: a physical process able to transmit a mark in a continuous way
2. Causal interaction: spatio-temporal intersection between two causal processes that modifies both
