Plantinga begins by examining the Gettier-type problems that internalist accounts of knowledge face. Having shown these difficulties, Plantinga is now able to set the stage for his externalist approach to warrant. This he does by explaining our design function: Any well formed human being who is in an epistemically congenial environment and whose intellectual faculties are in good working order will typically take for granted at least three things: that she has existed for some time, that she has had many thoughts and feelings, and that she is not a thought or feeling (Plantinga 50).
He then examines three apparent weak points of externalism and show not only are they strong points, only a fool would challenge them: memory, other persons, and testimony. In the nature of the case we do not have basic beliefs about these three entities in the sense that evidentialism and classic foundationalism require (especially memory and testimony; solipsism has a host of problems beyond this). Throughout this defense we see the vindication of Thomas Reid.
The book is quite difficult and technical, though. The sections on probability will lose all but the most formidable philosophers. While reading these chapters one is reminded of Eowyn’s comments to Merry before the battle: “Courage, Merry; it will soon be over.”
He then gives a (mostly) wonderfully lucid discussion of coherentism, classic foundationalism, and Reidian foundationalism. Coherentism sees truth as a source of warrant in the existing relations of one’s beliefs: does a belief “cohere” and “mesh” in a larger noetic structure? Plantinga suggests this is inadequate because coherentism only tells us of the doxastic relationships between beliefs. Warrant, by contrast, needs far more, experience among other things (179). Classical foundationalism is wrong because it is self-referentially incoherent. It is not the case that the foundationalist claim (a belief is properly basic because it is either self-evident to me or immediately present to my senses) meets its own criteria: it is not self-evidently true nor is it available to the senses (182). This leaves us with Plantinga’s position: Reidian foundationalism. If a belief is formed in proper circumstances according to its proper cognitive design, it has warrant.
The book began well and ended well. The middle sections were good, too, but likely only of interest to the most doughty of analytic philosophers. While I agree with Plantinga’s thesis, there are some shortcomings (but these can be excused because they have been treated in later works). The section on Reidian foundationalism, for example, while fundamentally sound, seemed to lack, forgive the pun, coherence in articulation. I kept seeing what RF was not in relation to classical foundationalism, but very little on what it was. The final chapters on naturalism are interesting, but have since been further refined in Plantinga’s later works.