Step 1: Extract the thesis of the passage.
The passage insists that probability has meaning only for a class (e.g., “all insured men aged 41 in a given country not engaged in dangerous occupations”). For a single person (a non-replicable one-off), “probability of death” has no meaning.
Step 2: Test each statement against the thesis.
(1) Singular, non-replicable events can be assigned probability.
This contradicts the passage: the author explicitly denies attaching probability to a single, unique case. Reject 1.
(2) Probability requires class data.
This is exactly the author’s requirement—probability is defined over a class with countable outcomes/frequencies. Accept 2.
(3) Class data predict any specific future event.
The passage does not license determinative prediction for a particular case; it only allows a frequency statement about the class. Reject 3.
Step 3: Conclude.
Only Statement 2 follows. \[ \boxed{\text{2 only (B)}} \]