Revision:As philosophy - theory of knowledge - knowledge and justification

What is knowledge?

Tripartite Definition: Justified True Belief

What is it?

  • In order know y:
    • x must believe y
    • x must be sufficiently justified in believing y
    • Y must be true
  • All of above are considered necessary and sufficient to claim knowledge
  • I.E. If I believe that pigs can fly, and I have video evidence that can justify this; it does not mean I know pigs can fly because it is not true.
  • I.E.2 If I believe that Fred Smith will win the lottery because my Dog told me so in a dream, and then he does I cannot claim to ‘know’; I only had true belief.



  • Conditions may not be necessary:
    • Belief – can we claim knowledge without belief? Knowledge is epistemological while belief is psychological – surely suggests some disparity?
      • E.g. a mother loves/ believes in teenager very much, despite them being unruly/ reckless. After 16 years of them reeking havoc – if confronted we would not allow her the excuse that ‘she didn’t know’ what child was up to, only that she refused to accept/ believe it.
    • Justification – can have knowledge without being able to provide justification (only true belief). How do we know when it is justified?
      • E.g. how can we justify that we are in pain? Or what if we forgot being taught something?
    • True – subjective POV suggests that knowledge is simply an accepted belief. Therefore, truth is not necessary for knowledge. (there are no objective truths)
      • For example, people once ‘knew’ that earth is flat.
  • Conditions may not be sufficient
    • Gettier style example: Fred drives a very rare car – only a few in existence. Whenever he is in college, he parks it in car park. You see this rare car in the car park and assume he is in college, which he is. However, it turns out that his broke down and he took the train – the rare car, coincidently, was somebody else’s. Did you know they were in college?



  • More often than not if somebody fulfils these conditions they will have knowledge
  • May be philosophically incorrect, but often works practically and is simple



What is it?

  • Responds to Gettier examples (where justification has become detached from subject/ accidental knowledge)
  • In order for p to know x:
    • P must believe x
    • X must be true
    • X must have been reached by a reliable method.
  • All of above are necessary and sufficient for knowledge
  • Reliability definitions:
    • Correctness: does it consistently produce correct results?
    • Causality: a method is reliable if the knowledge is brought about by X. E.g. if your pen is on the floor and you claim it rolled off the table and it did then this caused your belief; you do know. If somebody pushed it though, then the pen rolling off the table did not cause the belief. (answers gettier example)



  • Notion of reliability – what is this? How do we know if something has been obtained by reliable method? This is vague and unsatisfactory.
    • Correctness: this leaves us with very few reliable methods. All justifications/ investigations are open to error/ doubt! (Wittgenstein examples). Also, how often does it have to produce correct results to be reliable? Who defines this?
    • Causality: can know things without them being caused by X.
      e.g.Universal truths: I know that humans make tinned spam. Is this caused by the fact that humans make spam or do I simply know it logically?
    • Causality 2: how do we know that something is causal? Requires god-like observer; not practical
    • Causality 3: may not be sufficient:
      E.g. Orange in the box: a box with oranges lights up when something is placed inside. It does this when anything is placed inside. Somebody puts an orange inside and another (gullable) person sees the box and a sign saying ‘lights up when object inside’. They assume the presence of an orange in the box. Orange caused belief, but do they know it is there? After all, something else could be inside.
  • Belief – not always necessary (see above)



  • Bi-passes Gettier example
  • Solves problems with justification
  • Simple

Justifying Knowledge

Justification and evidence

Three methods of justification:

  • Predictive and explanatory power: a belief is justified if it provides greater predictive and explanatory content of phenomena than other possible propositions.
    E.g. Galileo’s suggestion that Earth not at centre of universe allowed for greater prediction/ explanation than other theories and was therefore justified
  • Acquisition by reliable method: a belief is justified when it has been obtained by a reliable method.
    E.g. consulting an expert on the metabolic pathway of aerobic respiration would be a more reliable method than asking your granny (assuming she’s not an expert).
  • Reasonableness/ Probability: a justified belief is one that can be considered reasonable (not absurd) or has a high probability of being true
    E.g. it is reasonable to suggest that ‘it will rain in Britain in the next year’, due to statistical probability. Therefore, this is justified.



  • In order to be justified, a belief must appeal to another equally or more certain/ basic belief.
  • In other words, each beliefs rests on another – like a chain of justification
  • Sometimes deductive, but beliefs may be inferred from others.
  • Nearly always coupled with idea of indubitable foundational belief – one that is self justifying and certain or ultimately basic. (to counter infinite regress)
  • Thus ideas exist in two categories – those that are foundational and those that are inferred



  • Infinite regress – belief can never be justified because the chain must end; can’t have infinite beliefs to justify with. Is always possible to ask ‘but what justifies that’. Leads to scepticism. (answered above)
  • Foundational beliefs – do these exist? Certainty does not (and cannot) include the impossibility of doubt unless analytic a priori (rationalist) which is not linked to the world and therefore cannot be used as justification!
    • Sense data as foundational (empiricist) – phenomenalism-type examples:
      • the distinction between sense-data and sense perception – false only know sense data through relation with world; not naming sense data but relating it to all the coloured objects (not foundational).
      • + might be mistaken about even sense data; e.g. might genuinely thought you saw blue instead of purple.
  • Inference – what is this? How can a foundationalist give an account of this? Surely many opposing beliefs could be inferred from a more basic one?



  • An instinctual method of justification
  • Simple
  • Infinite regress can be answered (very few flaws)



  • In order to be justified, beliefs must fit together as one holistic system.
  • Beliefs must cohere with each other without contradiction
  • More circular/ web-like justification
  • Beliefs do not have to be basic – they just have to fit
  • (Our beliefs must fit with our observations also)



  • Closed systems – our coherent sets of beliefs form a closed system of which there may be many; e.g. religion and science. But surely there can only be one set of true coherent beliefs about the world?
  • Non-coherent knowledge:
    • might want to claim knowledge on things that do not need to cohere (e.g. ‘it is raining now’ – know this because can see it, not because it coheres)
    • or non-coherent knowledge (e.g. something odd that does not fit with your beliefs).


Strengths/ Counters

  • Scientific Theories/ Philosophy of Science – fits with accepted beliefs of how science works such as paradigm shifts etc. Science is simply a coherent system which fits with our observations of the world and is justified as such. It behaves like a coherent system.
  • Closed systems – argument makes assumption that closed systems will be equally viable. However, surely there are certain observations/ more basic beliefs which systems must fit with? Some will be better. Universal Translator Example
  • Non-coherent knowledge:
    • Surely this belief still coheres with other beliefs such as ‘I can trust my sense perceptions
    • You wouldn’t claim you still knew something that didn’t fit – you would either claim you were mistaken or reconsider your system. (science examples + paradigms)
  • Explains relativism – shows us why two people may disagree while knowing all the observations. People appeal to facts within their own coherent system.

Acquiring Knowledge and Types of Knowledge


  • Knowledge can be obtained through reason alone
  • There are innate ideas which we might obtain knowledge from
  • (Often:) Knowledge = deductive/ foundationalist – Descartes.
  • Sceptical of the senses/ uncertain a posteriori knowledge
  • Real knowledge = complete certainty/ is necessary
  • 2 + 2 = 4
  • Kant – synthetic a priori = non trivial: held that we placed forms upon the world in order to interpret it and that these forms are a priori.



  • Triviality – if all knowledge is analytic a priori = all inconsequential
  • Logic/ disconnection - reason does not have to be connected to world, rather we check logic/ reason with observations
  • Closed Systems – multiple possible logical deductions/ inferences; e.g. many possible scientific paradigms. Cannot be reason alone because only one world! Seems to suggest that logic cannot be wrong.
  • Innate ideas – see foundationalism
  • Investigation ­– why else would this be necessary if knowledge = reason
  • Kant – even if are forms, must still investigate in order to obtain them. Very different universe would = very different forms.



  • Geometry ­­- obviously can obtain knowledge through reason
  • Certainty – cannot be contingent as logical/ not based on outside world. (although the foundations could be wrong! E.g. if space was ever so slightly curved, geometry would have to alter definition of straight line for more accurate calculation).
  • Understanding – is obvious that knowledge cannot come from senses alone. E.g. WAX example from Mr. Descartes.



  • Knowledge is obtain(ed/able) through senses alone
  • Knowledge IS contingent
  • Ideas/concepts = copies of sensations
  • Reject innate ideas
  • Like a posteriori knowledge
  • Sense data are certain/ basic



  • Understanding/ concepts – even accepting knowledge requires investigation; how do we understand this? Surely we need reason to do so. E.g. causality
  • Sense data certainty – see phen./ foundationalism
  • concepts = copies – surely if true the we should have concept of a particular red when considering. But, can consider many.
  • Parasitic language – e.g. can’t sense ‘on’ or ‘underneath’.
  • How do they account for a priori knowledge?
  • How do we gain knowledge of time?



  • Utility – necessarily linked to the world and how it functions.
  • Investigation – we have a need to investigate
  • Sensational knowledge – certain knowledge (such as I am hungry) is obtained through sense perceptions.



  • Theory stating knowledge is impossible – there is only varying degrees of belief
  • Uses philosophical doubt – which seeks to doubt that which is doubtable (rather than that which you have reason to doubt = ordinary doubt)
  • Everything is open to doubt, and therefore nothing is certain = no knowledge (see Theories of Perception, 5.)



  • Impracticality – cannot maintain this stance in everyday life; it is therefore useless
  • Paradox – how can you claim that you are certain nothing is certain
  • Indubitable statements – e.g. the cogito or ‘language exists’ or ‘you can think logically’
  • Wittgenstein + doubt – doubt always requires certainty. We cannot doubt one thing unless we are certain of another. E.g. Descartes never doubts his logicality.
  • Elastic word limits/ ordinary language ­– knowledge ≠ impossibility of doubt. To claim otherwise is to stretch the word beyond its ordinary use.



  • Allows testing of knowledge/ assumptions



  • Analytic – tautological
  • Synthetic – non tautological
  • A priori – without (before) investigation
  • A posteriori – with investigation
  • Contingent – could be wrong
  • Necessary – could not be wrong
  • Tripartite definition ­– theory of knowledge defining knowledge as justified true belief
  • Reliabilism – theory of knowledge defining knowledge as true belief obtained by a reliable method
  • Foundationalism – theory stating that in order to be justified a belief must be supported by another more certain/ basic one.
  • Coherentism – theory stating that in order to be justified a belief must cohere or fit with our already accepted beliefs.
  • Empiricism - knowledge is obtained through senses alone
  • Rationalism - knowledge is obtained through reason alone.
  • Innate knowledge – ideas which we are born with and do not have to learn
  • Scepticism – the belief that knowledge is ultimately impossible – there are only varying degrees of belief