one can apply either to avoid this computation or to drastically simplify it; you will see
several examples of these tricks later in the book.
Finally, it is worthwhile to compare the probabilities of the object being animate before
(Equation (2.6) versus

2.4.1
Bayes rule
Bayes rule (also called Bayes theorem) is simply the expression of a conditional probability in terms of the converse conditional probability and the two relevant unconditional
probabilities:
P (A|B ) =
P (B |A)P (A)
P (B )
(2.4)
Bayes ru

usage) the phrases principal and interest and interest and principal ; if we abbreviate these
two orderings as p and i (we might denote the union of the two orderings as cfw_interest,
principal ) then mathematically our linguist is saying that P (p) = P (

Chapter 2
Univariate Probability
This chapter briey introduces the fundamentals of univariate probability theory, density
estimation, and evaluation of estimated probability densities.
2.1
What are probabilities, and what do they have to
do with language?