E MMON B AC H

T HE AL GE BRA OF E VE NT S
0. I NT R ODUC T I ON
A numbe r of wri t ers have c omme nt e d on t he cl ose paral l el s bet ween t he
mas s - count di st i nct i on in nomi nal syst ems and t he aspect ual cl assi fi cat i on
of ver bal expr essi ons (Al l en, 1966; Tayl or , 1977; Mour el at os, 1978; L.
Car l son, 1981; Hoe pe l ma n and Rohr er , 1980) t hat has be e n t he subj ect
of muc h at t ent i on in r e c e nt year s in linguistics and phi l osophy. To t ake
j ust one class of exampl es f or now, t her e is a paral l el bet ween t he t wo
sets of di st i nct i ons in t hei r c ooc c ur r e nc e pat t er ns wi t h expr essi ons
denot i ng number s or amount s, as in Exampl es ( l a) - ( 4b) :
(1)(a) Muc h mud was in evi dence.
( b) ( *) Much dog was in evi dence.
(2)(a) J ohn sl ept a l ot last ni ght .
(b)(*)John f ound a uni cor n a l ot last ni ght .
(3)(a) Many dogs wer e in t he yard.
( b) ( *) Many muds wer e on t he fl oor.
(4)(a) J ohn fell asl eep t hr ee t i mes dur i ng t he ni ght .
(b)(*)John sl ept t hr ee t i mes last ni ght .
(By t he use of " ( *) " I i nt end t o i ndi cat e t wo things: t hat we have t o do a
cer t ai n a mount of wor k t o i mpose a speci al i nt er pr et at i on on t h e sen-
t ence and t hat t he i nt er pr et at i on is shaped by t he pr es ence of t he numbe r
or quant i t y expressi on. )
Th e basic ai m of this paper is t o t r y t o el uci dat e this pr opor t i on:
event s: pr ocesses: : things: stuff. Th e a c c ount draws heavi l y on a r ecent
paper by Gode ha r d Li nk on t he count - mass- pl ur al domai n (Link, 1983)
as well as on t he wor k of a numbe r of wri t ers who have cont r i but ed a
gr eat deal t o our under st andi ng of "ver b- cl assi f i cat i on". 1 In Sect i on 1, I
r evi ew bri efl y t he cl assi fi cat i on and in Sect i on 2 Li nk' s analysis f or t he
nomi nal domai n. In Sect i on 3, I set f or t h our pr oposal s about event s and
pr ocesses and in Sect i on 4 t ake up a numbe r of pr obl ems, some with,
some wi t hout , sol ut i ons.
Linguistics and Philosophy 9 (1986) 5-16.
@ 1986 by D. Reidel Publishing Company
6 EMMON BACH
1. EVENTS, PROCESSES, STATES
Here' s a scheme of the kinds of distinctions we want to deal with (based
on L. Carlson, 1981, but using our terminology in part):
st at es
dynami c (o] st at i c ( b)
eventualities
non-states
processes ( c) events
protracted ( d) momentaneous
happenings ( e) cutrninations ( f )
Typical examples are:
(a) sit, stand, lie + LOC
(b) be drunk, be in New York, own x, love x, resemble x
(c) walk, push a cart, be mean (Agentive)
(d) build x, walk to Boston
(e) recognize, notice, flash once
(f) die, reach the top
I will take it as given that it is necessary to have at least this much of a
classification if we are to deal adequately with the syntax and semantics
of English. A great deal of evidence for this point has been given in the
last several years, for example in connection with attempts to understand
the English progressive and similar constructions in other languages. 2
Most recently, Hans Kamp (1981) and E. Hinrichs (1981) have shown
the necessity for these distinctions for interpreting narrative structures.
2. MASS, COUNT, AND PLURAL IN THE NOMI NAL SYSTEM
In the work alluded to above, G. Link (1983) argues for the adoption of a
somewhat more richly structured model than those made available, for
example, in Montague' s work. 3 In this section, I will briefly sketch the
outlines of Link' s system.
The main idea in Link' s semantics is to give more structure to the
domain of individuals. Along with ordinary individuals like John and
Mary as in standard interpretations of the predicate calculus or in
THE ALGEBRA OF EVENTS 7
Mont ague' s work we are to have pl ural i ndi vi dual s like t hose denot ed by
the children or John and Mary as well as quant i t i es of " s t uf f " or mat t er
t hat corresponds to i ndi vi dual s of bot h kinds, such as t he gol d in Ter r y' s
ring or t he stuff t hat makes up t he pl ural i ndi vi dual J ohn and Mary. 4
Mor eover , cert ai n rel at i ons among t hese vari ous subdomai ns and the
el ement s maki ng t hem up are proposed. I pr esent t he essentials in an
i nformal way (for precise details t he r eader is r ef er r ed to Li nk, 1983).
St art wi t h a set A~ of i ndi vi dual s of t he mor e fami l i ar sort, for exampl e,
John, Mar y, this t abl e, Ter r y' s ring. We ext end this domai n by means of a
j oi n oper at i on to defi ne a superset E as follows:
(i)
(ii)
& ~_Ei
If a, fl ~ Ei t hen the / -j oi n (individual join: aUifl) of ol and
/3~E.
So t he / -j oi n of J ohn and Mar y is in E~ if each of J ohn and Mar y is. We
establish a partial orderi ng on t he member s of E~ (-~) by sayi ng t hat a is
"l ess t han or equal t o" (or "i s an i ndi vi dual part (/ -part ) of ") 13 j ust in
case the / -j oi n of ot and 13 is j us t / 3 itself. Thus t he i ndi vi dual John is an
/ -part of t he plural i ndi vi dual s John and Mar y or Ter r y' s ring and John.
The i ndi vi dual s f r om whi ch we st art ed are at oms in t he big st ruct ure t hat
we are building.
Among t he el ement s of A~ (and hence Ei) t here is a subset whi ch
forms a special subsyst em of its own. These are t he port i ons of mat t er or
stuff, for exampl e, t he gol d of whi ch Ter r y' s ri ng is composed. Thi s
subsyst em has its own j oi n and partial or der i ng (m-j oi n: Urn; m-part :
--<,1). Call this set D~. Fi nal l y, we need to specify the rel at i onshi p
bet ween t he syst em of D~ and t he rest of the domai n. We do this by
assumi ng a mappi ng hi- f r om i ndi vi dual s (at omi c and plural) to the stuff
out of whi ch t hey are composed. Thi s mappi ng shoul d satisfy t he
r equi r ement t hat t he or der i ng ---~ among t he i ndi vi dual s be pr eser ved in
t he or der i ng -<m among t he quant i t i es of mat t er maki ng t hem up (it is a
homomor phi sm) . Mor eover , b4-(x) = x j ust in case x c Di. For exampl e, if
John is an / -part of t he plural i ndi vi dual Ter r y' s ring and John, t hen t he
stuff maki ng up John had bet t er be an m- par t of t he stuff maki ng up
Ter r y' s ri ng and John. Not e t hat we have two di fferent part -whol e
relations. John is an / -part of t he i ndi vi dual John and Mary, but John' s
arm is not an i ndi vi dual part of John, bot h are at oms. On t he ot her hand,
t he stuff maki ng up John' s ar m is an m- par t of t he stuff maki ng up John.
Not e f ur t her t hat t he same quant i t y of stuff can cor r espond to many
di f f er ent individuals. For exampl e, t here may be an i ndi vi dual falling i nt o
8 EMMON BACH
the extension of the singular count noun man, say John, but there is also
a plural individual falling under the extension of the plural noun cells
such that the values for hi. given the two arguments are identical. The
two individuals are members of the equi val ence class induced by the
relation of material identity. Link calls a system of this sort a "Bool ean
model structure with homogeneous kernel " (boosk).
Some consequences of Link' s const ruct i on that I find interesting and
apposite for the present cont ext are these (I haven' t given enough details
to show that these consequences follow):
(5) Suppose Hengst a is a horse and Hengi st is a horse. Then the
plural individual Hengst a and Hengi st is not a horse, but is in
the extension of horses (contrast mass terms).
(6) Suppose the plural individual A and B is in the extension of
horses and likewise C and D. Then the plural individual A, B,
C, and D is also in the extension of horses (cf. mass terms).
(7) Even if the individual that is the quantity of gold composing
Terry' s ring is old, Terry' s ring need not be.
(8) The two meanings of sentences like John and Mary lifted the
box (each vs. together) can be nicely represent ed in Link' s
semantics by adding the interpretation provi ded for the plural
individual to the interpretation provi ded, say, in Mont ague' s
PTQ.
3. T HE ALGEBRA OF EVENTS AND PROCESSES
We now want to try out Link' s ideas in the domai n of eventualities, that
is, to characterize the structure of the model when we ext end it to the
domain of event s and processes, which for the moment I will consider
just as new kinds of elements in the (sorted) domain. I will start by
considering event s to be analogous to the singular and plural individuals
and bounded processes (' bits of process' ) analogous to the portions of
mat t er that make up the ' material extensions' of those individuals.
Our new system will then include the following:
(1) Ee: the set of event s with join operations Ue and partial
ordering -<e (a compl et e atomic Bool ean algebra);
(2) Ae ~ Ae: atomic event s;
(3) De _ Ae: bits of process with join I t e and partial ordering -<p
(a compl et e join semilattice);
THE ALGEBRA OF EVENTS 9
(4) In addi t i on, we will need two t empor al rel at i ons on Ee X Ee:
~: "st ri ct l y pr ecedes " (tr., irr., asymm. ),
: " over l aps " (nont r. , refl., symm. ) (cf. Bach, 1981; Kamp,
1980);
(5) a homomor phi s m he f r om (Ee, U e, <e, ~, ) to
(De, Up, --<v, ~, ) such t hat
(i) he(a) = a iff a c De,
(ii) he(a Ue/3) = he(a) Uphe([3), and
(iii) aRl 3~he( a) R' he( [ 3) for R =- < e, oc , o and R' = _< p, oc,
respect i vel y.
For purposes of illustration, I will assume t hat tenseless cl auses of
Engl i sh are t o be i nt er pr et ed as denot i ng sets of event ual i t i es, i.e.
member s of t he domai n Ee (for some di scussi on of t he general ki nd of
model st r uct ur e I assume, see Bach, f or t hcomi ng) . So here are some
exampl es of t he kinds of event ual i t i es t hat cor r espond t o t he above
distinctions:
(9) John kiss Mary: at omi c event
(10) Mar y st umbl e and Mar y twist her ankl e: pl ural event
(11) Mar y st umbl e: at omi c event
(12) Peopl e di scover t he hi dden cove: pl ural event
(13) Sally build a cabin: at omi c event
(14) Sally pound in a nail: at omi c event
(15) Jones poi son t he popul ace: at omi c event
(16) Jones pour poi son i nt o t he wat er mai n: at omi c event
Our homomor phi s m h ( hencef or t h I will drop subscripts on all symbols
wher e it is cl ear f r om cont ext whi ch domai n we are consi deri ng) will
del i ver up for us t he bounded bits of process cor r espondi ng t o i nst ances
of each of t hese event types. Just as in t he case of the nomi nal domai n it
is exceedi ngl y difficult to find Engl i sh expressions whi ch cor r espond to
t hese ' pure processes' (cf. our r emar ks on "(*)" aft er our first examples).
Some i nt ui t i ons I want to capt ur e wi t h r egar d to t he above exampl es are
these:
A d (10) and (11): a pl ural event of t ype (10) has (necessarily) a
si ngul ar event of t ype (11) as a n/ - pa r t , and t he processes associ at ed by h
wi t h t he l at t er is a p- par t of t he process associ at ed wi t h t he former.
A d (13) and (14): an event of t ype (14) mi ght ver y well be such t hat its
process is a p-part of t he process associ at ed wi t h an event of t ype (13).
1 0 E MMON B AC H
Ad (15) and (16): Events of these two types might be materially
(processually) equivalent while the events themselves are different. Thus,
Jones might very well intentionally pour poison into the water main (in
order to rid waterbeds of bedfish) and not intentionally poison the
populace (cf. Davidson, 1980, passim).
Just as in the nominal domain (Link, 1983), I will assume that our
interpretation assigns various predicates to different classes according as
they fall under the sort of classification outlined above. (HOW we decide
or do this will not be my concern in this paper.) So Dying names an
atomic kind of event, Running doesn't, and so on. Familiar properties of
these various kinds of eventualities will follow, such as indivisibility and
additivity (cf. L. Carlson, 1981; Bach, 1981): no proper p-part of a dying
is a dying; the fusion of two runnings is a running, but no two dyings are
a dying. I will return below to some interesting problems connected with
such facts.
4. S OME P AR AL L E L S AND P UZ Z L E S
We have found it quite instructive to think about parallels and differen-
ces obtaining between the two domains. In a number of places questions
and observations about one of the domains has led us to consider prob-
lems in the other domain in a new light.
4.1. Packaging and Grinding
It has frequently been observed that practically any count noun or name 5
can be used as a mass term: There was dog splattered all over the road;
Much missionary was eaten at the festival (David Lewis's Universal
Grinder, cf. Pelletier, 1979). Moreover, the opposite switch occurs as
well: muds = 'kinds of mud' , ice-cream = 'portions of ice-cream' (Uni-
versal Packager). In each case, we have a change of meaning with no
overt marking in the form of the word.
In the verbal domain, we find the same sort of phenomenon. Dowty
(1972) observed that practically any process verb can be used ' event-
ually', given the right context.
One of his examples was the process verb look for. One of the
characteristics of process verbs is that they don' t occur comfortably in
the context NP finished Ving: ?I finishes looking for a unicorn. Yet in the
context of a library with a well-defined search procedure, a sentence like
I finished looking for a book seems perfectly ordinary.
T HE AL GE B R A OF E VE NT S 11
In Engl i sh, t he way of swi t chi ng back and f or t h bet ween count and
mass, e ve nt and pr ocess t ypi cal l y i nvol ves no change in t he f or ms
i nvol ved. Th e di f f er ence is r at her i nduced by t he cont ext . In ot her
l anguages, ove r t mor phol ogi cal pr ocesses or rel at i onshi ps are avai l abl e
or obl i gat or y, f or exampl e, in t he per f ect i ve- i mper f ect i ve cont r ast s in
Slavic l anguages. Thi s raises i mpor t ant quest i ons of pr i nci pl e f or t he
analysis of Engl i sh. Do we want t o i nvoke f or mat i on rul es wi t h zer o-
mor phol ogy (i dent i t y oper at i ons in t he synt ax of words), as in Li nk' s rul e
f or f or mi ng t he mass- t er m count er par t t o a c ount noun like apple? Or do
we want t o s omehow gi ve meani ngs f or wor ds t hat are unspeci f i ed al ong
this di mensi on?
It seems t o me t hat t her e is an as ymmet r y in t hese r el at i ons bet ween
count and non- c ount meani ngs t hat runs in t he same di r ect i on in t he t wo
domai ns. Tha t is, if we st art wi t h a count meani ng and der i ve t he
non- c ount meani ng (as in Li nk' s rul e) t her e seems t o be a r egul ar and
pr edi ct abl e meani ng. Th e mas s - t er m apple seems t o mean t he stuff such
t hat t her e is at l east one appl e such t hat t hat stuff st ands in t he
const i t ut i on- r el at i on t o t he appl e (but see bel ow, Sect i on 4. 3 f or some
r emai ni ng pr obl ems wi t h this account ) . On t he ot her hand, goi ng in t he
ot her di r ect i on, t he connect i on seems muc h less syst emat i c, as al r eady
not ed. A be e r may be a ser vi ng of beer or a ki nd of beer . Similarly, in t he
ver bal domai n, when we put a pr ocess expr essi on i nt o a count cont ext ,
we must c ome up with some ki nd of cor r es pondi ng event , but j ust w.hat it
is is r el at i vel y f r ee, per haps t he begi nni ng of t he pr ocess in quest i on, or
some bounde d por t i on of it. Thi s as ymmet r y is pr edi ct ed by our f or mal
set -up: t her e is a f unct i on ( homomor phi s m) f r om t he count el ement s t o
t he non- c ount ones, but it is a ma ny- t o- one mappi ng so t hat we can' t in
gener al expect a uni que answer when we ask what count el ement this
por t i on of non- c ount stuff mi ght cor r es pond t o.
Count el ement s c ome as al r eady bounde d and di scr et e items. The r e -
f or e we can count t hem. Non- c ount el ement s don' t and t her ef or e need
some addi t i onal speci fi cat i on in or der to be used as count abl e expr essi ons
wi t h pl ural s or number s. Fur t her , expr essi ons whi ch car ve out measur es
or quant i t i es of st uf f , - pounds of, por t i ons of, et c. - cannot go wi t h pur e
count - i t ems in t he si ngul ar, but de ma nd i nt er pr et at i on of t he count - i t em
as mass- t er m or pr ocess count er par t . Mor e ove r , f or pl ural s size and
measur e are r el evant t o det er mi ni ng nat ur al ness and useful ness of t he
par t i cul ar expressi ons; two tons of horses is odd f or pr act i cal r easons in a
way t hat two tons of beans or fifty tons of horses ar e not (cf. L. Car l son,
1981, on t hese and ma ny ot her details). Th e r e ar e i nt er est i ng puzzl es
about count i ng t hat we will r et ur n t o bel ow ( Sect i on 4. 4).
12 EMMON BACH
4.2. The Partitive Puzzle
Dowt y (1978) and ot hers have di scussed t he socal l ed ' i mperfect i ve
par adox' (I prefer t o call it a puzzle). Briefly, t he puzzle is this: how can
we char act er i ze t he meani ng of a progressi ve sent ences like (17) on t he
basis of t he meani ng of a simple sent ence like (18) when (17) can be t rue
of a hi st ory wi t hout (18) ever bei ng t rue?
(17) John was crossi ng t he street.
(18) John crossed t he street.
(See Vl ach, 1981; Dowt y, 1979.)
Nat ural l y, we want t o use t he apparat us we have set up t o pr ovi de an
account of t he Engl i sh progressi ve, perhaps al ong t he lines of Vl ach
(1981). Thi nki ng about how t o do this has led us to see t hat t here is a
per f ect l y paral l el pr obl em in t he nomi nal domai n, whi ch we call t he
' part i t i ve puzzle' .
Consi der Li nk' s account of t he fol l owi ng sent ence:
(19) Ther e is apple in t he salad.
Li nk' s i nt er pr et at i on amount s to this: t here are some apples, such t hat
some of t he stuff maki ng t hem up is pr esent in t he salid. Not e the
existential quant i fi cat i on over apples; t he sent ence coul d not be t rue of a
hi st ory whi ch never had any appl es in it. Thi s seems reasonabl e enough
for this sent ence, but consi der t he fol l owi ng:
(20) Thi s is part of a paper on nat ural l anguage met aphysi cs.
(21) We f ound par t of a Roma n aqueduct .
It seems as if (20) coul d be t rue even t hough (alas!) t he paper in quest i on
never r eached ful fi l ment , and (21) t rue when t here no l onger is an
aqueduct or even if progress on t he const r uct i on was i nt er r upt ed f or ever
by hor des of barbari ans f r om the nort h.
Let us l ook mor e closely at Li nk' s account . The denot at i on of t he mass
t er m cor r espondent mp of a pr edi cat e P is gi ven as this (p. 309):
~rnp~ : = {x e D I x -- sup [h~e~]}.
(Here, sup st ands for supremum.) That is, t he denot at i on of apple (used
as a pr edi cat i ve mass t erm) is t he set of quant i t i es of mat t er t hat are
m- par t s of t he val ue of h appl i ed t o t he set of apples in t he world. Thus,
no apples, no apple. But t here coul d surel y be a worl d in whi ch it was
possible to artificially manuf act ur e appl e wi t hout t here bei ng any apples,
or for less f ar f et ched exampl es, consi der agai n Exampl es (4) and (5).
T HE AL GE B R A OF E VE NT S 13
Such exampl es show t hat we need t o allow f or a mor e i ndi r ect r el at i on
bet ween t he denot at i on of a mass pr edi cat i ve mass t er m and t he cor -
r espondi ng count pr edi cat e. Basi cal l y, we need t o be abl e t o say when
cer t ai n stuff is of t he r i ght ki nd t o qual i fy as falling under t he ext ensi on
of t he mass t er m, or bet t er we need t o assume t hat we can say when this
is t he case. To act ual l y gi ve cr i t er i a is no par t of linguistics (cf. Put nam,
passim).
Fur t her , al t hough we have assumed t hat t he t wo domai ns of t hi ngs and
stuff ar e separ at e, it seems t o me t o be r eas onabl e t o assume t hat our
knowl edge of what qualifies as a quant i t y of appl e or mud or gol d is
based on our under st andi ng of what is meant by t he t er m phr ases apples,
mud, or gold, under s t ood as names f or ki nds (G. Car l son, 1977) or
pr oper t i es ( Chi er chi a, 1982). Bot h Car l son and Chi er chi a ar gue t hat
such t er ms ar e mor e i nt ensi onal t han t he pr oper t i es of Mont ague, whi ch
ar e f unct i ons f r om wor l d- t i me pai rs to i ndi vi dual s. We may say t hat a
pr ope r t y or ki nd det er mi nes such a f unct i on, whi ch t hen may be used t o
get t he denot at i ons of t he cor r es pondi ng pr edi cat i ves apple, gold and
mud. So t o say t hat t her e is appl e in t he salad is t o say t hat t her e is some
stuff in t he salad of t he r i ght sor t as t o qual i fy as appl e and t he l at t er
i nvol ves appeal i ng t o our knowl edge of t he ki nd of appl es or t he
pr oper t y of bei ng an appl e. It shoul d t hen fall out of our t heor y t hat
par t i cul ar appl es are made of appl e and so on.
4.3. How Old Is the Gold?
Li nk pr ovi des a ni ce analysis of t he puzzl e pr es ent ed by a s ent ence like
this:
(21) Th e gol d maki ng up Ter r y' s ri ng is ol d but t he ring itself is
new (not old).
Puzzl es like this one ar e among t he best evi dence f or not i dent i fyi ng
t hi ngs wi t h t hei r mat er i al count er par t s . But t her e is still a pr obl em.
Th e i nt er pr et at i on of (22) is this: Th e x such t hat x makes up Te r r y' s
ri ng and is gol d is ol d but Te r r y' s ring is not old. No cont r adi ct i on, si nce
x and Ter r y' s ring ar e not t he same t hi ng, x is j ust t he val ue of h wi t h
Te r r y' s ri ng as ar gument . But now consi der a s ent ence like (23):
(23) Th e snow maki ng up this s nowman is qui t e new but t he H20
maki ng it up is ve r y ol d (and t he H and O e ve n ol der!).
Th e i nt er pr et at i on of this s ent ence comes out like this: Th e x such t hat x
1 4 E MMON B AC H
const i t ut es t he s nowman and x is snow is new but t he y such t hat y
const i t ut es t he s nowman and y is wat er is ve r y ol d (not new). Thi s is
cont r adi ct or y accor di ng t o Li nk' s a c c ount si nce x and y must be i den-
t i cal (since h is a f unct i on) . If we fol l ow Li nk' s sage advi ce - " o u r gui de
in ont ol ogi cal mat t er s has t o be l anguage i t sel f " (Li nk, 1983, pp. 303f. ) -
t hen it seems t o me t hat we have t o set t hi ngs up in such a way t hat we
can r ef er t o i ndi vi dual s under a descr i pt i on, somehow. Thi s puzzl e is
cl osel y c onne c t e d t o t he next one.
If we fol l ow Li nk' s advi ce, t hen we must acknowl edge t hat t wo t hi ngs
wi t h cont r adi ct or y pr oper t i es cannot be i dent i cal . Thus t he snow maki ng
up t he s nowman and t he H2 0 maki ng it up must be di f f er ent , and nei t her
can be e qua t e d wi t h t he undi f f er ent i at ed quant i t y of mat t er gi ven by
Li nk' s homomor phi s m. Wha t t o do?
Th e first possi bi l i t y is t o acknowl edge t hat our l anguage allows us t o
t al k about chai ns of composi t i on, so t o speak. Th e snow in t he s nowman
is itself ma de up of t he wat er in t he s nowman (plus air) and so on. What
t he exampl e shows, t hen, is t hat we cannot use t he const i t ut i on r el at i on
di r ect l y in an i nt er pr et at i on of a phr ase like the snow making up the
snowman. Our i nt er pr et at i on of such phr ases must be such t hat it does
not hol d t hat if x makes up a and y makes up a t hen x = y. We can
essent i al l y keep all of Li nk' s appar at us i ncl udi ng t he homomor phi s m and
t he equi val ence classes ge ne r a t e d by it but mer el y amend t he way in
whi ch Engl i sh wor ds like make up, constitute or phr ases like the gold in
the ring are i nt er pr et ed.
A s econd way woul d be t o r e move al t oget her t he ent i t i es in D f r om
t he domai n of i ndi vi dual s. I expl or e one way of doi ng this based on
Cresswel l ' s (1973) met aphysi cs of possi bl e worl ds and i ndi vi dual s in
Bach, f or t hcomi ng. Thi s woul d a mount t o sayi ng somet hi ng l i ke this:
Stoff an sich (j ust l i ke t he Ding an sich) can have no pr oper t i es, at l east
as f ar as our l anguage is concer ned.
4.4. How Many Things Are There in the Room?
In bot h domai ns ma ny wri t ers have poi nt ed t o char act er i st i c pr oper t i es
l i ke addi t i vi t y, subdivisibility, ant i addi t i vi t y and ant i subdi vi si bi l i t y whi ch
pl ay cl ear rol es in gi vi ng an a c c ount of ent ai l ment r el at i ons among
sent ences: Hengi st ' s ear can' t be a hor se, mud pl us mud is mud, a hor se
pl us a hor se i sn' t a hor se, and so on. But in bot h domai ns t her e ar e cl ear
and or di nar y exampl es of count i t ems t hat don' t fol l ow t hese rest ri ct i ons.
The s e ar e wor ds like thing, event, happen, and so on. Suppose it is t r ue
t hat somet hi ng happened, t hen in t he nor mal case t her e ar e smal l er
THE AL GE BRA OF EVENTS ] 5
s ube ve nt s t ha t ma k e up t he bi g t hi ng t ha t h a p p e n e d t ha t ar e al so
ha ppe ni ngs . Si mi l ar l y f or t hi ngs . I n b o t h d o ma i n s we ar e at s o me t h i n g of
a l oss t o t r y t o a ns we r que s t i ons l i ke t hes e:
(24) Ho w ma n y t hi ngs ar e t he r e i n t he r o o m?
(25) Ho w ma n y e v e n t s t o o k pl a c e i n t he l ast h o u r ?
Gu p t a ( 1980) has s t r es s ed t he i mp o r t a n c e of cr i t er i a of r e i de nt i f i c a t i on
a nd i ndi vi dua t i on i n t he l ogi c of c o mmo n nouns . Ou r di s cus s i on he r e
s hows t ha t pr i nc i pl e s of i ndi vi dua t i on ar e cr uci al f or expr es s i ons a nd
c o n c e p t s i n t he v e r b a l d o ma i n as wel l . We f ol l ow Li n k i n not r e qui r i ng
t ha t t he s u b d o ma i n s Di a n d De be a t o mi c Bo o l e a n al gebr as . Thi s is as it
s houl d be. I t is n o t p a r t of l i ngui st i cs t o d e c i d e wh e t h e r all ma t t e r is
a t o mi c or all h a p p e n i n g s ar e r e duc i bl e t o l i t t l e g r a n u l e s of pr oc e s s .
I n d e e d , if c o n t e mp o r a r y phys i c a l t he or i e s ar e t o be be l i e ve d, s uc h
ul t i ma t e que s t i ons ar e bas i cal l y i n c o h e r e n t . Ev e n t s a n d pr oc e s s e s ar e
di sj oi nt , a nd t hi s s e e ms t o be mo r e an ar t i f act of o u r l a n g u a g e or
c o n c e p t u a l i z a t i o n s of t he wo r l d t ha n s o me t h i n g a b o u t t he wor l d i t sel f, so
t ha t p r o b a b l y h e r e t o o o u r s t r i ct l y s e ma n t i c t he or i e s s houl d r e ma i n si l ent .
AC KNOWL E DGE ME NT S
A l a r ge p a r t of t he s u b s t a n c e of t hi s p a p e r de r i ve s f r o m j oi nt wo r k wi t h
Ba r b a r a Pa r t e e , first p r e s e n t e d by us at a c o l l o q u i u m at St a n f o r d Un i -
ver s i t y, h e n c e t he f r e q u e n t " we " . I t a ke ful l r es pons i bi l i t y f or e r r or s o r
i nf el i ci t i es. Gr a t e f u l a c k n o wl e d g me n t is t e n d e r e d t o t he Ma x Pl a n c k
I ns t i t ut fi i r Ps yc hol i ngui s t i k i n Ni j me g e n , wh i c h s u p p o r t e d p a r t of t he
wo r k r e p o r t e d on her e.
NOT E S
Besides the writers mentioned in the first paragraph, see for example, the excellent survey
in Dowry (1979) and the references cited there. The classic modern works dealing with
verb-classification are Kenny (1963), and Vendler (1957). To my knowledge, Verkuyl
(1972) was the first extensive work which recognized the importance of these distinctions
for linguistic theory.
2 Dowty (1977), Vlach (1981), for example.
4 It is important to notice that Link's proposals differ crucially from previous attempts to
deal with plurals which constructed interpretations for plurals in purely set-theoretic ways,
as for example in Bennett's (1974) classic treatment.
4 Strictly speaking, we will have the corresponding NP denotations, that is, property sets
for the individuals mentioned. I ignore this complication throughout the paper.
s Link does not deal with the use of names in mass contexts: They put five pounds of Porky
into the stew.
16 EMMON BACH
REFERENCES
Allen, R. L.: 1966, The Verb System of Present-Day English, Mouton, The Hague.
Bach, Emmon: 1981, ' On Time, Tense, and Aspect: An Essay in English Metaphysics', in
Peter Cole (ed.), Radical Pragmatics, Academic Press, New York, pp. 62-81.
Bach, Emmon: forthcoming, 'The Metaphysics of Natural Language', to appear in Pro-
ceedings of the Seventh International Congress on Logic, Methodology, and Philosophy of
Science.
Carlson, Gregory: 1977, Reference to Kinds in English, Ph.D. dissertation, University of
Massachusetts, Amherst.
Carlson, Lauri: 1981, 'Aspect and Quantification', Syntax and Semantics 14.
Chierchia, Gennaro: 1982, 'Nominalizations and Montague Grammar: A Semantics
Without Types for Natural Language', Linguistics and Philosophy 5, 303-354.
Cresswell, Max J.: 1973, 'Logics and Languages', Methuen, London.
Davidson, Donald: 1980, Essays on Actions and Events, Clarendon Press, Oxford.
Dowty, David R.: 1972, Studies in the Logic of Verb Aspect and Time Reference in English,
unpublished Ph.D. dissertation, The University of Texas, Austin.
Dowty, David R.: 1977, 'Toward a Semantic Analysis of Verb Aspect and the English
'Imperfective' Progressive', Linguistics and Philosophy 1, 45-79.
Dowty, David R.: 1979, Word Meaning and Montague Grammar, Reidel, Dordrecht.
Gupta, Anil: 1980, The Logic of Common Nouns, Yale University, New Haven.
Hinrichs, Erhard: 1981, Temporale Anaphora im Englischen, University of Tuebingen,
Magisterarbiet.
Hoepelman, J. and C. Rohrer: 1980, ' On the Mass Count Distinction and the French
Imparfait and Passe Simple', in C. Rohrer (ed.), Time, Tense and Aspect, Tuebingen,
Niemeyer, pp. 629-645.
Kamp, Hans: 1980, 'Some Remarks on the Logic of Change', Part I, in Christian Rohrer
(ed.), Time, Tense, and Quantifiers, Niymeyer, Tuebingen.
Karnp, Hans: 1981, 'l~v6nements, repr6sentations diseursives et r6f6rence temporelle',
Language 64, 39-64.
Kenny, Anthony: 1963, Action, Emotion, and Will, Humanities Press, New York.
Link, Godehard: 1983, 'The Logical Analysis of Plurals and Mass Terms', in R. Bauerle,
Ch. Schwarze, and A. von Stechow (eds.), Meaning, Use, and Interpretation of Language,
pp. 302-323.
Mourelatos, Alexander, P. D.: 1978, 'Events, Processes, and States', Linguistics and
Philosophy 2, 415-434.
Pelletier, F. J.: 1975, 'Non-Singular Reference', in F. J. Pelletie~ (ed.), Mass Terms: Some
Philosophical Problems, Reidel, Dordrecht, pp. 1-14.
Putnam, Hilary: 1975, Mind, Language, and Reality, Philosophical Papers, Vol. 2,
Cambridge University Press, Cambridge.
Vendler, Zeno: 1957, 'Verbs and Times', Philosophical Review 56, 143-160.
Verkuyl, Henk: 1972, On the Compositional Nature of the Aspects, Reidel, Dordrecht.
Vlach, Frank: 1981, 'The Semantics of the Progressive', Syntax and Semantics 14, 271-292.
De p a r t me n t o f Li ngui s t i cs
Uni versi t y of Mas s ac hus e t t s Amh e r s t
Amh e r s t , Mas s ac hus e t t s 0 1 0 0 3 , U. S . A .