What if the measures (variable values) don't match?
What if two variables that really are different (conceptually) have the same name?
In this version the log does not warn you that you've shot yourself in the foot.
DATA SURVEY2;
INPUT SUBJ GENDER :$1. wEIGHT hEIGHT YEAR;
... identical ...
But the output shows some signs of problems.

More Perverse than Example 2.2 in C&P

SUBJ    GENDER    HEIGHT    WEIGHT    YEAR

  1       M          68       155      91
  7       F          72       205      90
  9       M          66       120      93
  5       F         102        63      92
  8       M         250        70      91

There is no "do what I meant, not what I say". Variable with same names are treated as one.