By Hannes Leitgeb

In distinction to the present culture in epistemology, the focal point during this e-book is on low-level inferences, i.e., these inferences that we're frequently now not consciously conscious of and that we proportion with the cat within reach which infers that the fowl which she sees choosing grains from the dust, is ready to fly. most likely, such inferences are usually not generated through particular logical reasoning, yet logical tools can be utilized to explain and examine such inferences.

Part 1 offers a in simple terms system-theoretic explication of trust and inference. half 2 provides a reliabilist conception of justification for inference, with a qualitative suggestion of reliability being hired. half three recollects and extends a number of platforms of deductive and nonmonotonic common sense and thereby explains the semantics of absolute and excessive reliability. partially four it truly is confirmed that qualitative neural networks may be able to draw justified deductive and nonmonotonic inferences at the foundation of dispensed representations. this can be derived from a soundness/completeness theorem with reference to cognitive semantics of nonmonotonic reasoning. The appendix extends the idea either logically and ontologically, and relates it to A. Goldman's reliability account of justified trust.