The Global Standard Stratotype-section and Point (GSSP) concept is the mode
rn standard for defining chronostratigraphic units in the geological time s
cale, and over the past decade the working groups involved in locating and
characterizing GSSPs have rekindled appreciation for stratigraphy within th
e profession. The rules adopted hy the ICS for establishing the GSSP are, h
owever, less restrictive than, and to some extent in conflict with, the pri
nciples set forth by the international Subcommission on Stratigraphic Class
ification (ISSC) under the leader-ship of Hollis Hedberg (1976). Under the
ICS rules, global boundaries may he established ad hoc for units at any lev
el without regard for the constraints of nested chronostratigraphic hierarc
hy, even while acknowledging that the stage supposedly holds the critical p
lace. Furthermore, the ICS rules, in advocating that GSSPs be located stric
tly according to their correlatibility, have the effect of requiring precon
ceived boundaries that cor respond to global geohistorical events. These ap
parently practical shortcuts have serious consequences against which the He
dherg guidelines specifically, warned. Not only does the disregard of stage
unit- and boundary-stratotypes do more or less violence to the existing li
terature, but event-based definitions are inherently unstable in stratigrap
hic space. We review here the problems associated with the definition of a
GSSP, using the controversy over the Paleocene/ Eocene boundary as a case h
istory.