The Relation of Grammar to Cognition--a Synopsis 
Leonard Talmy 
Program in Cognitive Science / Center for Human Information Processing / UC San Diego 
Abstract 
A sentence (or other portion of discourse) is 
taken to evoke in the listener a meaning complex, 
here called a "cognitive representation". The lex- 
ical elements of the sentence, to simplify, by and 
large specify the content of the cognitive represen- 
tation, while the grammatical elements specify its 
structure. Thus, looking systematically at the 
actual notions specified by grammatical elements can 
give us a handle for ascertaining the very makeup of 
(l~nguistic-) cognitive structuring. We accordingly 
examine a number of grammatically specified notions, 
observe the categories and systems in which they 
pattern, and speculate on broader cognitive connec- 
tions. 
Some provisional findings have already emerged. 
Grammatical specifications for structure are prepon- 
derantly relativistic or topological, and exclude the 
fixed or metrically Euclidean. The categories in 
which grammatical notions pattern include: 
plexity perspectival mode 
state of boundedness level of synthesis 
state of dividedness level of exemplarity 
degree of extensionality axial characteristics 
pattern of distribution scene-breakup " 
Grammatical specification of structuring appears to 
be the same, in certain abstract characteristics, as 
the structuring of visual perception. 
O. Introduction 
A sentence (or other portion of discourse) is 
taken to evoke in the listener a particular kind of 
experiential complex--here to be termed a "cognitive 
representation" or "CR".I There appears to be a sig- 
nificant way in which different portions of the lan- 
guage input specify, or code for, different portions 
of the CR. The major finding is that--for a first 
approximation--the lexical fraction of a sentence 
codes mainly for the content, or substance, of a CR, 
while the grammatical fraction of a sentence codes 
mainly for the structure of a CR. Determining the 
structure within a realm of phenomena has been a cen- 
tral concern for analytic science, including linguis- 
tics and psychology. With grammar seen in the above 
light, it can be used in determining the structure, 
of the language-related portion of human cognition, 
with possible connections to further portions. In 
particular, Iookingsystematically at the actual not- 
ions specified by grammatical elements can give us a 
handle for ascertaining the xery nakeup of (linguis- 
tic-) cognitive structuring. ~ The beginnings of such 
an endeavor are the aims of this paper. 
Several ideas here require some immediate elab- 
oration. The distinction between lexical and gram- 
matical is made entirely formally--i.e., without any 
reference to meaning--on the basis of the distinc- 
tion between open-class and closed-class. 3 All open- 
class elements--i.e., the stems of nouns, verbs, and 
adjectives4--are considered lexical. Everything else 
is considered grammatical. Included here are all 
closed-class morphemes and words--inflections, par- 
ticles, adpositons, conjunctions, demonstratives, 
etc.--as well as syntactic constructions, grammatical 
relations~ categorial identities, word order, and 
intonation. Terminologically here, "grammatical 
element" will be used to refer to any of these. 
The nature of content and of structure, and the 
distinction between them, are not understood well 
enough to be addressed analytically in this paper and 
must be left to our intuitive sense of the matter.5 
Taking them for granted, however, we can now more 
finely characterize the linguistic-cognitive cross- 
relationships noted earlier. While most of a CR's 
content is specified by the lexical fraction of a 
sentence, the lexical items do usually specify some 
structural notions along with the contentful ones. 
The grammatical elements of a sentence more unalloy- 
edly specify only structural notions;~and specify them 
more determinately in the case of conflict with a 
lexical item, establishing perhaps the majority of a 
CR's structure. 6 
In other work in the present direction--notably 
Fillmore's (e.g., 1975, 1976)--concern has also been 
with ascertaining structre, but the sentence elements 
used as starting-points have generally been lexical 
items with prominently inmixed structural specifica- 
tions (like buy and sell)~ The present work, in part 
a complement to the other, takes advantage of gram- 
mar's greater directness and completeness:in speci- 
fying structure~ 
This paper is divided into three sections. In 
the first, a sampling of grammatical elements is ex- 
amined for the notions that they specify, both as an 
introduction to out method and for the aim of notic- 
ing properties common to such notions as well as pro- 
perties excluded from them. In the second, we pre- 
sent a number of the categories in which grammatically 
specified notions have been observed to pattern. In 
the third, we speculate on broader cognitve connec- 
tions. 
14 
1. The Nature of Grammatically Specified Notions 
In this section we examine a small sampling of 
grammatical elements for the particular component 
notions that they specify. The sample will give a 
heuristic indication of the kinds of notions that get 
grammatically specified as well as of kinds of no- 
tions that possibly never do. The excluded kinds 
will be seen as readily specifiable by lexical ele- 
ments. A further comparison between the character- 
istics of grammatically specified notions and of 
lexically specified ones is then made. To indicate 
the major finding at the outset, it seems that gram- 
matical specifications for structure are preponder- 
antly relativistic or topological, and exclude the 
fixed or metrically Euclidean. 
For a first simple case, many languages have in- 
flections for the noun (English has -B and -s) 
that specify the uniplex or the multipTex instantia- 
tion of the object specified by the noun. By con- 
trast, no languages appear to have inflections that 
specify the redness or blueness, etc.--i.e., the par- 
ticular color--of the object specified by a noun. 
In the preceding, the underlined are instances of 
"notions". The first set are grammatically specified 
a~d can be readily seen to play a structuring role 
in a CR. 7 The second set are perhaps never found 
specified by grammatical elements, though they are 
everywhere found specified by lexical elements \[such 
as (red and blue). 
For another case we consider a deictic like the 
English this or that as in This chair is broken. A 
grammatical element of this type specifie~ the loca- 
tion of an indicated object as being, in effect, on 
the speaker-side or the non-speaker-side of a concep- 
tual partition drawn through space (or time or other 
qualitative dimension). This integral specification 
can be analyzed as containing the following component 
notions (enclosed by quotes): 
(I) 
a-b. a 'partition' that divides a space into 
'regions'/'sides' 
c-e. the 'locatedness' (a particular relation) of a 
'point' (or object idealizable as a point) 
'within' a region 
f-g. (a side that is the) 'same' as or 'different' 
from 
h-i. a 'currently indicated' object and a 'currently 
communicating' entity 
Notions that might at first be ascribed to such deicq 
tics, such as of distance or perhaps size, prove not 
to be, on the evidence of sentence-pairs like (2): 
(2) a. This speck is smaller than that speck. 
b. This planet is smaller than that planet. 
The CRs evoked by (2a) and (b) differ greatly, in- 
volving tiny objects millimeters apart or huge objects 
parsecs apart. Yet the sentences differ only lexic- 
ally, not grammatically. Hence, the CRs' notions as 
to the magnitude of size or distance cannot be traced 
to the deictics (or to other grammatical elements) in 
the sentences. Thus, the notional specifications of 
a this or a that appear, in part, to be genuinely 
topological: the establishment of a partition remains 
a constant, but its position can vary unlimitedly (or, 
using topology's characterizability as "rubber-sheet 
geometry", the partition's distance away can be 
stretched indefinitely) without any constraints im- 
posed by the deictics' specifications per se. This 
finding about the deictics alerts us to noticing 
whether any grammatical elements make specifications 
about magnitude. A spot check through English and 
various other languages suggests that--while there are 
apparentl~ grammatical specifications for relative 
magnitudeS--there are possibly never any for absolute 
or quantified magnitude, whether of size, distance, 
or other parameters. 
For a third case, we consider the type of adposi- 
tion that specifies, for a moving object, certain 
characteristics of path and of point- or frame-of- 
reference. An example of this type is English through 
as used, e.g., in: 
(3) a. I walked through the water. 
b. I walked through the timeber (i.e., woods). 
In this usage, through specifies, broadly, 'motion 
along a line that is within a medium'. The component 
notions contained here include: 
(4) 
a-e. 
f. 
g. 
h-i. 
'motion'--i.e., 'one-to-one correspondences' 
between 'adjacent' points of 'space' and 
adjacent points of 'time' 
motion that describes a 'line' 
the locatedness of a line within a 'medium' 
a medium, i.e., a region of three-dimensional 
space set apart by the locatedness within it 
of 'material' that is in a 'pattern of dis- 
tribution', of a certain range of character 
(still to be determined) 
Again, with (3a) and (b) differing only lexically, any 
notional differences in their CRs cannot be attributed 
to through. Thus, not within the specificational 
purvue of that element are: the 'kind nf substance' 
comprising the medium and the 'sensorimotor character- 
istics' attendant on executing the motion--as, here, 
those attendant on wading vs. weaving amidst obstacles. 
With other sentence pairs like 
(5) a/b. I crawled/ran through the timber. 
(6) a/b, I zig-zagged/arced throught the timber. 
it can be further determined that 'rate of motion' and 
'shape/contour of linear path' are also not specified 
by the grammatical element. 
As one step in a program to ascertain any proper- 
ties common to grammatically specified notions, the 
notions just found are gathered together in Table I. 
For heuristic purposes, the notions are very provis- 
ionally divided into three groups on the basis of 
their relation to topology. In group (a) are the 
notions that properly belong, or are readily definable, 
in the actual mathematical system of topology. In 
group (b), the notions might not be part of topology 
proper but intuitively seem like those that are--and 
might be includable in a related mathematical system 
that could be constructed. In group (c) are the no- 
tions that fall outside of any usual conception of a 
mathematical system. The number of notions in the 
first two groups combined is 13, while the third has 
6--an indication of a preponderant propensity for 
grammatical elements to specify quasi-topological no- 
tions. The ratio in this direction is in fact im- 
proved if we consider that even several notions in 
group (c)--the bottom three--resemble topological ones 
in the sense of involving relativistic relationships 
between quantities rather than absolutely fixed 
quantities. 
15 
(7) Table 1: Some notions found to be specified 
by grammatical elements 
a. topological b. topology-like 
partition 
region/side 
point 
line 
locatedness 
within 
uniplexity 
multiplexity 
one-to-one 
correspondences 
same 
different 
pattern of distribution 
"adjacency" of points 
(monotonicity) 
c. non-topological 
matter 
space 
time 
motion 
medium 
currently indicated/ 
communicating entity 
For a complementary program of ascertaining any 
properties excluded from grammatical specification, 
the notions found above not'~to be specified by the 
elements investigated are listed in Table 2. Rather 
than topological, topology-like, or relativistic, 
these notions involve Euclidean-geometric concepts 
(e.g., set distance,:size, contour), quantified mea- 
sure, and various particularities of a quantity--in 
sum, characteristics that are absolute or fixed. 
(8) Table 2: Some notions seemingly never specified 
grammatically 
absolute/quantified magnitude kind of substance 
(of distance, size, etc.) speed 
shape/contour of line color 
sensorimotor characteristics 
The provisional conclusion to be drawn from these 
findings is that, if grammatical specifications largely 
correspond to (linguistic-) cognitive structuring, then 
the nature of that structuring is largely relativ- 
istic or topological rather than fixed or absolute. 
In a search for contrasts between grammatical and 
lexical specification, a difference that presents 
itself at this point is that the relativism vs. abso- 
lutism restrictions do not apply to the latter. Lex- 
ical items can specify topological and relativistic 
concepts, as the very words listed in Table 1 attest 
to. And they can also specify Euclidean or absolute 
concepts. Thus, for the notion of color in Table 2, 
there are such lexical items as red, b\]ue; for con ~ 
tour, there are circle, straight; for quantified 
magnitude, there are inch, mile; for sensorimotor 
characteristics, there are wade, nimble, effort. 
For a further contrast between the grammatical 
and the lexical type of specification, we consider 
the full complement of both element-types in a single 
whole sentence, viz., that selected in (9): 
(9) A rustler lassoed the steers. 
We first list the grammatical elements p~esent in the 
sentence~and the notions that they specify: 
(10) 
a. -edz 
b. the: 
c. a: 
'occurring at a time before that of 
the present communication' 
'has ready identifiability for the 
addressee' 
'not before in discussion or otherwise 
readily indentifiable for addressee' 
d. -s: 'multiplex object' 
e. a...-~: 'uniplex object' 
f. the grammatical category of "verb" for lasso: 
'eventhood' 
g/h. the gram. category of "noun" for rustler/steer: 
'objecthood' (one possible spec. of "N") 
i/j. the grammatical relations of "subject"/"object" 
for rustler/steer: 
~t~7~-patient ' (among possible specs.) 
k. active voice: 
'point-of-view at the agent' 
I. intonation, word-order, state of auKiliaries: 
'the speaker-"knows" the situation 
to be true and asserts it' 
The lexical items in the sentence can have their spec- 
ifications characterized as follows: 
(II) A complex of concepts involving: 
a. rustler: 
b. steer: 
c. lasso: 
property ownership, illegality, mode 
of activity 
appearance, physical makeup, relation 
to animal k~ngdom 
institution of breeding for intended 
purposes, esp. human consumption 
certain materials (a body and a lasso) 
in certain configurations 
movement sequences of materials' parts 
concomitant mental intentions, direc- 
tings, monitorings, etc. 
In surveying the lists, we can see these differ- 
ences emerge: The grammatical elements are more num- 
erous and their specifications seem simpler and more 
structural. Together, their specifications seem to 
determaine the main organizational and communicational 
delineations of the CR evoked by the sentence. The 
lexical elements are fewer in number, but their specG 
ifications are more complex and seem to comprise most 
of the content of the CR. The lexical specifications 
are complex in three ways: compared to a grammatical 
specification, each has a) more total information, 
b) greater intricacy of information, and c) more dif ~ 
ferent types of information together. 
These grammatical-lexical differences can be set 
into further relief by in turn varying one element- 
typewhile keeping the other constant. Thus, varying 
only the grammatical elements of (9), as is done in 
(12), seems to alter the organizational and communic- 
ational characteristics of the scene but to leave its 
basic contents intact: 
(12) Will the rustlers lasso a steer? 
Varying only (9)'s lexical elements, as in (13), shifts 
us to a new scene altogether, and yet the essential 
breakup of the scene and of the communicative setting 
seem to remain the same: 
(13) A machine cancelled the stamps. 
2. Categories of Grammatically Specified Notions 
The preceding sampling of grammatical elements 
has yielded a set of notions helpfultoward discovering 
common properties. But the set has been small and 
haphazardly arrived at. With a broader and more sys- 
tematic investigation, patterns of organization become 
evident. Grammatically specified notions can be seen 
to pattern in categories, and the categories,'in turn, 
16 
in integrated systems. In this section we look at 
some of these categories and systems. 
The grammatical elements here will not be treated 
in isolation, but in association with lexical items. 
That is, the grammatically specified structural no ~ 
tions will be considered in interaction with that 
portion of lexical specification that is also struc- 
tural. This interaction entails cognitive processing, 
and different cases of such processing will be con- 
sidered along the way. 
The note on methodology should be made that our 
direction of analysis has been from grammatical spec- 
ification to category, not the reverse. That is, the 
categories considered below were discovered to be 
relevant to the specifications of various grammatical 
elements. They were not part of some a priori concep- 
tual schema which then sought corroborative examples~ 
2.1 Dimension / Kind of Quantity 
The category of "dimension" has two member no- 
tions, 'space' and Itime'. The kind of "quantity" 
that exists in space is--in respectively continuous 
or discrete form--'matter' or 'objects'. The kind 
of quantity existing in time is 'action' or 'events' 
("action" is meant to refer to any obtaining circum- 
stance= not just (willed) motion). In tabular form, 
these notions relate thus: 
(13) space: matter/objects 
time: action/events 
A number of grammatical and lexical referents are 
specific with regard to one or the other pole of this 
category. But since the category cross-cuts the ones 
treated next, we will not exemplify it here but will 
endeavor in the following to present both space and 
time examples side by side. 
2.2 Plexity 
The category here to be termed "plexity" is a 
quantity's state of articulation into equivalent ele- 
ments. Where the quantity consists of only one such 
element, it is "uniplex", and where it consists of 
more than one, it is "multiplex". When the quantity 
involved is matter, plexity is, of course, equivalent 
to the traditional category of "number" with its com- 
ponent notions "singular" and "plural". But the pre- 
sent notions are intended to capture the generaliza- 
tion from matter over to action, which the traditional 
ones do not.9 
Specifications as to plexity are made by both 
lexical items and grammatical elements, and the in- 
terplay between the two when they are both in associa- 
tion must be noted. Example English lexical items 
that basically specify a uniplex referent are--for 
matter and action, respectively--bird and (to) sigh. 
They can occur with grammatical elements that them- 
selves specify a uniplexity, like those underlined 
in (14a) (many languages have here a more regular, 
overt system of markers than!:English). But they can 
also occur with grammatical elements that specify a 
multiplexity, as in (14b). In this association, such 
elements can be thought to trigger a particular cog- 
nitive operation--in this case, one of "multiplexing". 
By this operation, an original solo referent is, in 
effect, copied onto various points of space or time. 
(14) matter action 
a. uniplex A bird flew in. He sit (once). 
b. m~Itiplex Birds flew in. He kept sighing. 
The reverse of the preceding circumstances is 
also to be found in language. First, there are lex- 
ical items that intrinsically specify a multiplexity. 
English examples are furniture or timber (i.e., 'stan- 
ding trees') for matter and breathe for action, as 
used in (15a). And, too, there are grammatical ele- 
ments able to appear in association here, as in (15b), 
that signal an operation the reverse of multiplexing-- 
one that can be called '*unit-excerpting". By this 
operation, a single one of the specified equivalent 
units is taken and set in the foreground of attention. 
(15) 
a. multiplex 
b. 
matter 
action 
Furniture overturned in the 'quake. 
She breathed without pain. 
piece o_f_ffurniture overturned... 
She took a breath/breathed in... 
The grammatical elements that above signaled multi- 
plexing-- -s and keep -~--have a directly manifested 
surface form. The ones signaling unit-excerpting are 
in part abstract in form, as represented in (16): 
(16) matter action 
(a) Nunit of + Vdummy-~-)--~\[ \[ + X\]N 
eg: a piece of furniture take a breath 
or: + Prtcle (eg: in) 
2.3 State of Boundedness 
Another category of attributes specified both 
grammatically and lexically for a quantity is its 
"state of boundedness" When a quantity is specified 
as "unbounded", it is conceived as continuing on in- 
definitely with no necessary characteristic of finite- 
ness intrinsic to it. When a quantity is specified 
as "bounded", it is conceived as demarcated off as an 
individuated unit entity. 
Among English examples of lexical items, water 
and (to) sleep seem basically to specify unbounded 
quantities, whereas sea and (to) dress seem basically 
to specify bounded ones. These specifications are 
demonstrated by the words' respectively unacceptable 
and acceptable occurrence with the grammatical element 
in NPextent_of_t\]me , which specifies boundedness: 
(17) matter 
action 
a. unbounded *We flew over water in I hr. 
*She slept in 8 hrs. 
b. bounded We flew over a sea in I hr. 
She dressed in 8 mins. 
Now, there are grammatical elements suitable for 
co-occurrence with unbounded-type lexical items which 
therewith, in effect, trigger a cognitive operation 
of "bound@ng". By this operation, a portion of the 
specified unbounded quantity is demarcated and placed 
in the foreground of attention. Examples of such 
grammatical elements in English are: 
(18) matter (a) Nbounded_quantity of 
action for Nextent_of_time + 
Particular cases of them in use are: 
(19) We flew over a bodz of water in 1 hr. 
She slept for 8 hrs. 
17 
The question arises whether the reverse of the 
preceding circumstances is ever to be found in lan- 
guage. Entailed would be the existence of grammat- 
ical elements that, when used with lexical items 
specifying a bounded quantity, Would trigger an oper- 
ation of "debounding". By this, e.g., the referent 
of sea would be shifted to 'pelagic water', and that 
of-~tear, to take another lexical bounded case, 
would shift to 'lachrymal fluid'. It seems likely 
that such grammatical elements exist; the closest 
candidate known to the author is the French suffix 
-age, but this has a range of meanings and many oc- 
currence restrictions--and does not, e.g., happen to 
combine with the French words for "sea" or "tear".10 
2.4 State of Dividedness 
The category of "state of dividedness" refers to 
a quantity's internal consistency. A quantity is 
'~discrete" (or "particulate") if there are breaks in 
its oo~inuity. Otherwise, the quantity is "contin- 
uous". ~ Both lexical and grammatical elements are 
sensitive, in their specifications, to the distinc- 
tions of this category. But there appear to be no 
grammatical elements that solely specify discreteness 
or continuity for a quantity, and also none that sig- 
nal an operation for reversing a^quantity's lex~cally 
specified state of dividedness, zz In consequence, 
there is difficulty in demonstrating this category 
explicitly by itself, and so we defer its treatment 
until the next section, where it can be seen in in- 
teraction with the other categories. 
2.1 - 2.4 The Disposition of a Quantity 
The preceding four categories of attributes all 
pertain to a quantity simultaneously and, taken to- 
gether, can be considered to constitute a system of 
attributes that may be termed a quantity's "disposi- 
tion". The particular intersections of the several 
attributes will be the main object of attention here. 
These, firstly, can be schematized as in (19): 
(19) discrete continuous 
e ~ 6 ~,!/, ,,,%/ 
' ' 'illi/iIfI  , "" ", ,. r} " ' I. a, 7 : I ~:::unbounded 
multiplex < 
Q~... O <C-~ b°unded 
A B "~f 
uniplex .'- o ~ a 
+ the distinction between matter and action, 
which cross-cuts all of the above ~F 
Each intersection of attributes indicated here has 
been found specified by various lexical items. An 
example or two (most seen earlier) is given for each 
intersection in (20):14 
(20) A: timber/furniture B: water 
(to) breathe (to) sleep 
A: (a) family B: (a) sea/tear 
(to) button up (to) zip up 
a: (a) bird 
(to) sigh 
Now if the particular contentful referent for 
which one chooses a lexical item happens to be wedded, 
by that lexical item, to an unwanted set of structural 
specifications, there generally are grammatical means 
available for altering this to a desired set. Such 
means range in directness from specifying the single 
apt alteration to involving a circuitous sequence of 
operations. A number of starting- and ending-points 
for alterations, and the means for accomplishing them, 
are indicated in (21): 
(21) 
~--~A a stand of timber 
breathe for 1 hr. 
A-~,a a piece of furnit. 
take a breath/ 
breathe in 
B~B a body df water 
sleep for 1 hr. 
A --~a a member of a fmly 
go through a step 
of buttoning up 
A~A members of a fmly B-~B 
(A -*a -~A) 
button on and on 
a~A trees 
keep sighing 
tears (*tearage) 
(B -~a -~A -~-B) 
zip on and on 
a -~A a stand of trees 
(a -~A -~ A) 
sigh for a while 
2.5 Degree of Extensionality 
Implicit in the vertical dimension of the sche- 
matic arrangement in (19) is a further category I~ that 
can be called "degree of extensionality". This cate- 
gory has three member notions, terms for which are 
given in (22) together with schematics of the notions 
for the linear dimension: 
(22) point bounded extent unbounded extent 
Lexical items with either a matter or an action ref- 
erent can make concurrent structural specifications 
for their referent as to its basic degree of exten- 
sionality. Three examples--specifying objects of 
different linear extensionalities--are the words 
(23) speck ladder river 
Now a lexical referent that is perhaps most bas- 
ically to be conceived as of one particular degree of 
extensionality can, by various grammatical specifica- 
tions that induce a shift, be idealized as being of 
some other degree of extensionality. For a first ex- 
ample, consider the event referent of climb a ladder, 
which seems basically of bounded linear ~en-t-~of 
time), as is in fact manifested in (24) in conjunction 
with the grammatical element "in+ NPextent_of_time": 
(24) She climbed up the fire-ladder in 5 mins. 
With a different accompanying grammatical element, 
like the "at + NPpoint-of-time" in (25),(as well as 
different ~ntextual specifications), the event ref- 
erent of the preceding can be shifted toward idealiz- 
18 
ation as a point of time--i.e., as being point-dura- 
tional: 
(25) Moving along on the training course, 
she climbed the fire-ladder at exactly midday. 
This shift in the cognized extensionality of the ew 
ent can be thought to involve a cognitive process of 
"reduction" or of "taking the long-range view". The 
shift can also go in the other direction. The event 
referent can be idealized as an unbounded extent from 
the effect of grammatical elements like "keep -ing", 
"-er and -er", and "as + S", as in (26): 
(26) She kept climbing higher and higher up the 
fire-ladder as we watched. 
Here there would seem to have taken place a cognitive 
process of "magnification" or of "taking the close-up 
view". In such a process, a perspective is estab~ 
lished whereby the existence of any exterior bounds 
falls outside of view and attention--or, at most, are 
asymptotically approachable. 
The preceding event referent was continuous, 
but a discrete case can exhibit the same shifts of 
extensionality. One such case, perhaps to be con- 
sidered as most basically of bounded extent, is shown 
with that degree of extensionality in (27a). But the 
referent can also be idealized as a point, as in (27b) 
(it is clear that the cows here did not all die at the 
same moment, and yet the spread of their death times 
is conceptually collapsed into such a single moment). 
Or, the referent can be idealized as an unbounded ex- 
tent, as in (27c): 
(27) a. The cows all died in a month. 
b. When the cows all died, we sold our farm. 
c. The cows kept dying (and dying) 
until the serum finally arrived. 
The alternative idealizations of extensionality 
just seen as specifiable for an event referent are 
generally also available for an object referent. 
Thus, e.g., the referent of (a) box can be specified 
for idealization as a point or as a bounded extent 
(of area or volume). Some grammatical elements making 
such specifications~are illustrated in (28). Also set 
forth here are the homologies between these and the 
event-specific elements: 
(28) 
point The box is 20 ft. away from the wall. 
I read the book 20 yrs. ago. 
bounded extent The box is 2 ft. across. 
I read the book in 2 hrs. 
(point Within) The ball is in the box. 
bounded extent She arrived as I was reading the book. 
2.6 Pattern of Distribution 
The pattern of distribution of matter throughi: 
space or of action through time is a further category 
of notions that can be both grammatically and lexic- 
ally specified.16 For action through time--the only 
dimension we will be looking at now--this category 
together with the preceding one largely constitute 
the traditional category of "aspect". 
Several of the main patterns of distribution for 
action through time are shown schematically in (29) 
(the dots here, representing situatedness in comple- 
mentary states, should really be adjacent, but they 
are sketched apart with a connecting line to show the 
crossing of state-interfaces). Shown, too, are ex- 
ample verbs whose basic distributional specifications 
are as in the corresponding schematic: 
(29) 
one-way one-way full- steady- gradient 
non- resettable cycle state 
resettable ~ 
. .... . ,, 
die fall flash sleep widen 
carry 
One can determine that these lexical items have the 
specifications indicated by noting the grammatical 
elements with which they can and cannot occur (or, to 
put the latter case in our terms: ...grammatical ele- 
ments toward whose specifications they will not 
shift). A full demonstration is not in order here, 
but a few examples show the principle: The resettable 
type of a one-way event is distinguished from the 
non-resettable type by its compatibility in sentences 
like: He fell 3 times, which the other lacks: *He 
died 3 times. This same one-way form is distinguished 
from a full-cycle form by its ability to appear in 
sentences like: He fell and then got up, which the 
latter cannot do: *The beacon flashed and then went 
off. 
We can now consider the cirsumstance where a verb 
of one type appears with grammatical elements of an- 
other type and shifts in certain of its specifications 
of distribution. For an example we again take die, 
whose basic specifications can be adjudged as point- 
durational one-way non-resettable--schematizable, now 
more precisely, as: ~ . This verb is used with its 
basic specifications in a sentence like (30a). 
(30) a. He died as she looked on. 
b. He was (slowly) dying as she looked on. 
But in a sentence like (30b), the grammatical ele- 
ment "be+ -in__ng_" induces a shift. In effect, the 
infinitesimal interval between the two states involved 
for die--viz., 'aliveness' and 'deadness'--is spread 
out, with the creation thereby of an extent-durational 
gradient. This is the shift in the distribution pat- 
tern's structural type. But concomitantly, a shift 
in the basic contentful referent is engendered. In- 
stead of 'dying', the new gradient refers to 'mori- 
bundity'. The distinction becomes clear in noting 
that one can have been dying without having died, 
and, correlatively, one can have died without having 
been dying.17 
2.7 Perspectival Mode 
A specified action (which, in our terms, can as 
equally be static as involve change) has been seen to 
have its own, perhaps most basic, pattern of distri- 
bution through time. But, as it turns'out, there can 
be independent specification for a mode of attending 
to the action that has a distinct temporal pattern 
of distribution, one that is either equal or unequal 
to the action's. In what we shall now consider, 
there are two types of such "attentional" or "per- 
spectival mode" viz.: 
19 
(31) The assuming of: 
a. a steady-state long-range perspective point 
with synoptic scope of attention 
b, a moving close-up perspective point 
with local spope of attention 
To illustrate, we first consider an example with 
a basically steady-state referent, viz., objects in 
location. The (31a) type of perspectival mode--the 
one more congruent with such a referent--holds in 
(32a), multiply specified/determined there by the 
set of grammatical elements shown underlined. But 
by substituting grammatical elements coding for the 
(31b) perspectival mode, as is done in (32b), the 
scene evoked can be shifted to one where one's mental 
gaze or one's own projected location jumps in turn 
from object to object. In effect, a steady-state 
multiplexity of objects has been converted to a 
sequential multiplexity of events, viz., of concep- 
tualized encounters with the objects. 
(32) a. There are houses here and there in the valley. 
b. There is a house every now and then through 
the valley. 
In a comparable case, the moving-perspective form, 
shown in (33b), is the only mode that can be spec- 
i~fied using everyday language. One must resort to 
scientific language, as in (33a), in order to estab- 
ish the synoptic perspective: 
133) 
a. The telephone poles' heights form a gradient that 
correlates with their locations on the road. 
b. The telephone poles get taller the further down 
the road they are. 
The reverse of the preceding circumstances is 
also encountered. An example involving a sequential 
multiplexity of events is shown in (34a) with the more 
congruent moving-perspective mode specified. In (34b), 
the same referent instead becomes the object of syn- 
optic viewing. In metaphorical terms, the effect here 
is as if the vertical time line is tilted up into pre- 
sent-moment horizontality for integrated or summational 
assessment. 
(34) 
a. I took an aspirin time after time during/ 
in the course of the last hour. 
b. I have taken a number of aspirins in 
the last hour.18 
2.8 Level of Synthesis 
The category to be considered now pertains to 
bounded quantities, like those schematized in the 
A/B row in (19). One form of locution already seen 
to specify such quantities is the particular type of 
"NP of NP" construction illustrated in (35a). Here 
the second NP specifies the identity of the quantity 
involved, itself conceptualized as without intrinsic 
bounds, while the first NP specifies the bounding 
(or "portion--taking") per se of the quantity: 
(35) a. a set of trees 
b. a cluster of trees 
a body of water 
a puddle/drop of water 
Now, beyond the fact alone of bounding off a portion, 
the first NP can additionally specify the particular 
configuration or form that the portion takes, as in 
(35b). 19 Especially with regard to internally dis= 
crete quantities--as with a cluster of trees--the two 
NPs can here be seen as coding for two different 
"levels of synthesis": The later NP specifies an 
unsynthesized multiplexity, while the earlier NP spe- 
cifies a particular geatalt synthesized therefrom. 
There is a further cognitive distinction involved 
here that language usually makes: either level of 
synthesis can be placed in the foreground of attention 
while the other level is placed in the background. 
One grammatical form that specifies this involves 
placing the foregrounded NP-type first, as shown in" 
(36a)~ With the use of this grammatical device, 
moreover, predications can be made that pertain 
solely to one level of synthesis or the other, as 
seen in (36b)i 
(36) a. the cluster of trees / the trees in the cluster 
b. That cluster of trees is small. 
The trees in that cluster are small. 
There are certain surface forms, furthermore, whose 
referents are keyed to applying to only one or the 
other level of synthesis. Thus, together (toward 
each other) tends to correlate with multiple objects, 
while~_gn (upon itself) tends to correlate with a 
composite thereof: 
(37) The bricks in the pyramid came crashing 
together/?in. 
The pyramid of bricks came crashing 
in (upon itself)/?together. 
The preceding has involved shifting attention 
from a multiplexity to the gestalt that it consti- 
tutes. Also encountered in language are means for 
specifying the reverse: shifting attention from a 
gestalt to the components that constitute it. This 
procedure can take place when the starting lexical 
item specifies an entity taken ~o be already at the 
more synthetic level, as is the case with iceberg in 
(38a). By grammatical devices like those seen in 
(38b), such an entity can be broken down from con- 
ception as a coherent whole and presented in terms 
of component parts and their interrelations: 
(38) a. The iceberg broke in two. 
b. The two halves of the iceberg broke apart 
(*in two). 
Again we encounter a surface form--in two--that cor- 
relates with only one level of synthesis and not the 
other. 20 
2.9 Level of Exemplarity 
The specification for a multiplexity of objects 
can have a further cognitive distinction made per- 
taining to it. This distinction does notaffeCt the 
basic reference to all the members of the multiplex- 
ity, but addresses how attention is directed therein. 
Eithe~ the full complement of the multiplexity is in 
the foreground of attention, with perhaps individual 
items here and there singled out in the background 
of attention~ Or a single exemplar out of the multi- 
plexity is placed in the foreground of attention, 
with the remaining items more dimly conceived in the 
background of attention. Perhaps most languages have 
several grammatical devices for specifying this dis- 
tinction as to the "level of exemplarity". But Eng- 
lish stands out in the extensiveness of its forms: 
20 
there are different pairs of grammatical elements 
that mark the distinction for a number of distinct 
types of multiplexity. A rather full list of these 
pairs is illustrated in (39): 
(39) 
a. Oysters have siphons/a siphon. 41 
An oyster has siphons/a sipohon. L± 
b. All oysters have siphons/a siphon. 
Every oyster has siphons/a siphon. 
c. All the members raised their hand(s). 
d. Each member raised his hand(s). 
d. Many members raised their hand(s). 
Many a member raised his hand(s). 
e. Some members here and there raised their hand(s). 
A member here and there raised his hand(s). 
f. Members one after another raised their hand(s). 
One member after another raised his hand(s). 
g. Hardly any members raised their hand(s). 
Hardly a member raised his hand(s). 
h. No members raised their hand(s). 
No member (Not a member) raised his hand(s). 
i. She held a gun in both hands. 
She held a gun in either hand. 23 
2.10 Other Categories and Processes 
More notional categories and cognitive processes 
have been worked up than there is opportunity to pre- 
sent here. Some of this other material is treated 
in an earlier work, Talmy (1977) (which itself lacks 
some of the material presented here). But we will 
briefly indicate some of the concepts involved. 
The adjectives in a pair like sick/well behave 
differently in association with grammatical elements 
specifying vectoral degree, as shown in (40). In this 
they parallel the behavior of certain spatial expres- 
sions like at the border/past the border: 
(40) He's sllghtly~ "sick/past" the border.'~ 
~well/*at the border. 
(~well/at the border. 5 
He's almost~ ~ 
\[:sick/'past the border~ 
This behavior can be accounted for by positing that 
such adjectives are not simply "opposites", but, ra- 
ther, imply for some semantic notion, e.g., that of 
'health', a particular abstract topological axis of 
which each adjective labels a certain portion. The 
forms here seem in particular to imply a line bounded 
at one end; well refers to the end-point while sick 
refers to the remainder of the line. These are the 
lexical items' "axial characteristics", i.e., the 
particular (topological) relations ~ach has to a par- 
ticular semantic axis and to other items al~ng the 
same axis. Certain grammatical elements, like those 
underlined in (40), also specify axial characteris- 
tics. Used incompatibly, they can cause a shift in 
an associated adjective's specifications. Thus, in 
(41), sick seems to label an end-point, and of a 
different axis as well, that of 'feeling bad': 
(41) (After eating the shrimp, he felt worse and 
worse and) he was almost sick at one point/ 
he finally got sick in 5 hrs. 
Lexical expressions like cottage and hotel room 
may be taken to have "associated characteristics"-- 
here, respectively, those of 'permanent residence' 
and 'temporary lodging'. These attributes may mesh 
or conflict with the specifications of another ele- 
ment in the same sentence, e.g., with the directional 
adverb home, which specifies a permanent residence. 
In the cese of conflict, as in (42b), the lexical item 
is operated on by a cognitive process that leaves its 
essential characteristics intact but replaces its in- 
cidental characteristics: 
(42) a. He drove home to his cottage in the suburbs. 
b. He drove home to his hotel room. 
The "scene-breakup characteristics" of a lexical 
item like serve refer to its basic specification of 
a dyadic event, in particular, a social event invol- 
ving the two roles of 'host' and 'guest', as is mani- 
fested in (43a). But in a sentence like (43b), such 
a lexical item shifts to specifying a monadic event 
comparable to a basically monodic lexical expression 
like that in (43c). This shift in (42b) takes place 
in accommodation of the subject-plus-reflexive's 
single-role specification. (lhough this grammatical 
element is determinative in setting the role-number 
as monadic, the verb's influence remains: blended in 
here is the metaphoric suggestion of a dyad, as if 
both 'host' and 'guest' are to be found in the "I"): 
(43) a. The host served me some dessert from the kitchen. 
b. I served myself some dessert from the kitchen. 
c. I went and got some dessert from the kitchen. 
A major aim in cognitive linguistics must be to 
investigate the interactions between le~ical and 
grammatical specifications arising in a single sent- 
ence. Included here are the cognitive accommodations 
that take place where there are conflicting specifc- 
cations. A number of interactions have been provision- 
ally identified, and four seem definitely established: 
operations, shifts, blends (of two kinds: superimposed 
and introjected), and juxtapositions. The last three 
of these are treated at length in Talmy (1977). 
2.11 Nesting 
The operations and shifts seen in 2.1 - 2.6 need 
not take place singly. The output of one can serve 
as the input to another, up to as many as five hier- 
archical levels of l!nesting". While there are a num- 
ber of interesting examples of this for different 
types of matter and action, we will go directly to 
illustrating one of the longest cases; 
(44) 
a. The beacon flashed (as I glanced over). 
b. The beacon kept flashing. 
c. The beacon flashed 5 times in a row. 
d. The beacon kept flashing 5 times at a stretch. 
e. The beacon flashed 5 times at a stretch for 3 hrs. 
In (44a), the lexical verb flash appears with its 
basic structural specification as a point-durational 
full-cycle uniplex event. This undergoes the process 
of multiplexing, to yield the unbounded multiplexity 
in (44b). This then undergops bounding in (44c). 
This bounded multiplexity is then~first put through 
21 
the process of reduction to become idealized as a 
point, and this is in turn multiplexed, yielding 
(44d). This new unbounded multiplexity is finally 
then bounded in (44e). The nesting of structural 
specifications in this last stage can be represented 
schematically as in (45): 
3. Further Cognitive Connections 
Grammatically specified structuring appears to 
be similar, in certain of its characteristics and 
functions, to the structuring in other cognitive do- 
mains, notably that of visual perception. In parti- 
cular, the characteristic of being quasi-topological 
can be pointed to, and three major functions can be 
identified: classification, synoptics, and continuity. 
The thinking here is not equally far along on all i; 
these matters, but something of its directions can 
be indicated. 
Grammatical specifications can be seen to con- 
stitute a classification with regard to the vast var- 
iety of learned, conceived, and perceived material. 
They gather different portions of the material toge- 
ther into subdivisions distinct from each other. By 
this, any particular currently cognized element is 
associated with its implicit "subdivision-mates". 
An illustrative case here are the twenty-odd motion- 
related p~epositions in English, such as through and 
into, which together subdivide the domain of 'paths 
considered with respect to reference-objects' This 
domain covers a great and varied range, but any par s 
ticular "path" falls within the purvue of one or an - 
other preposition, associated there with other "paths". 
The associations are often language-specific and some- 
times seem arbitrary or idiosynchratic. Thus, as s~en 
earlier, classed together by through are such dissim- 
ilar cases as a straightforward liquid-parting course 
(walking through water) and a zig<zag obstacle-avoid- 
ing course (walking through timber). The question 
arises why such distinctions should be effaced by 
the grammatical system, while they are observed by 
the lexical and other cognitive systems. Why are 
grammatical elements--say, such prepostions--not a 
large and open class marking indefinitely many dis~i 
tinctions? One may speculate that the cognitive 
function of such classification lies in rendering 
contentful material manipulable--i.e., amenable to 
transmission, storage, and processing--and that its 
lack would render content an ineffective agglomeration. 
The original assumption made in this paper about 
grammatical specification involved the synoptic func- 
tion. That is, the grammatical elements of any par- 
ticular sentence together specify the structure of 
the cognitive representati6n evoked by that sentence. 
Their specifications act as a scaffolding or framework 
across which contentful material can be splayed or 
draped. It can be speculated that such structure is 
necessary for a disparate quanti~y of contentful mat- 
erial to cohere in any sensible way or to be simul- 
taneously cognized as a gestalt. 
In the course of discourse, a great welter of 
notions pass in rapid succession. But there are sev- 
eral ways in which a cognitive continuity is main- 
tained through this flux and a coherent gestalt is 
summated over time. For one, there are cognitive 
processes whereby the successive notions generally can 
be sensibly connected together or fit into a concep- 
tual matrix. For another, rhetorical specifications 
--all the yes, buts, on the other hands, and a num- 
ber of subtler elements not generally recognized for 
this--direct the illocutionary flow and make up the 
"logical" tissue of the discourse. Through this, gram- 
matical elements appear to play a determinative role. 
Their specifications establish a structural level with 
greater temporal constancy amidst more fleeting asp- 
ects of content. 
These forms of grammatically specified structuring 
seem to parallel forms discernable in the operation of 
visual perception. 24 First, the perception of any 
particular object is mediated by its association with 
related objects in a classificatory schema. 
Secondly, the welter of visual sensations cognized 
at any given moment for some whole scene is rendered 
coherent by the perception of structural delineations 
running through it. One specialized form of this is 
discernable when one intends to move through a space, 
say, from one to the opposite corner of a restaurant. 
The sensations of tables, chairs,etc, are, in effect, 
perceived in simplified spatial arrangements as if from 
an aerial view, and the plot of a course one could 
follow through that is sensed. 
Thirdly, in the course of motion through space 
over time, there is a great flux of visual sensations 
rushing past, but a sense of continuity is maintained 
by the perception of structure running through the 
successive scenes. Two levels of "scene-structure 
constancy" are maintained. In the first, the perceived 
delineations afford greater permanence than the sensory 
flux, but do slowly shift. This is the level where, 
say, in walking past a table, its perceived outline 
is maintained but shifts gradually from a quadrilateral 
to a trapezoid and back to a quadrilateral. A deeper 
level of greater constancy is also maintained, from 
which the table continues to be perceived as a rect~ 
angle no matter where one is in relation to it. For 
a final parallel with grammatical specification, the 
topology-like nature of visual perception is evident 
here. For certain abstract characteristics of a scene 
and its contents are maintained constant while other, 
more metrical and Euclidean characteristics are free 
to vary without relevance thereto. 
4. Notes 
1. The word "evoke" is used because the relationship 
is not direct. The CR is an emergent, compounded by 
~arious cognitive processes out of the sentence ele- 
ments' referential meanings, understanding of the pre- 
sent situation, general knowledge, etc. 
Our term "cognitive representation" is similar 
in purport to Fillmore's (1975)"scene" but is chosen 
over that more specifically visual term. The linguis- 
tically evoked somplex can have much from other sense 
modalities (notably som/kinesthetic and auditory) as 
well as meta-modal aspects. 
2. Comprehension, rather than production, is the dir- 
ection we limit ourselves to inthe initial endeavor. 
This direction would seem to yield more immediately 
reliable findings, since its starting point is with 
more overtly manifest, hence handleable, forms like 
grammatical elements rather than with meanings and 
experiential complexes, which rely more on introspec- 
tion and reports of introspection. Nevertheless, each 
direction does involve both the manifest and the ex- 
periential sides of language. 
3. This is a classical linguistic distinction. A 
class in which morphemes are formally gathered is con- 
sidered open if it is quite large and easily augment- 
22 
able relative to other classes. A class is considered 
closed if it is relatively small and fixed in member- 
ship. 
4. Also includable here are "lexical complexes" like 
lodge a complaint or zero in on. Excluded are adverbs, 
which seem in all languages to derive from the other 
three open classes rather than from any open class 
of specifically adverbial stems. 
5. Since the term "structure" has broad usage, we 
can help focus in on the intended sense with alter- 
native terms: "principles of organization", "pattern 
of delineations", "schematic framework". 
6. The fact of dual lexical specifications that can 
lead to conflict is a mojor issue that will be treated 
below under shifts. Some grammatical elements also 
cross the line and makecontentful specifications along 
with structural ones. This is a more tangential issue 
that can be touched on here. The crossing ranges from 
the incorporation of a single contentful notion to the 
orderly interweaving of contentful and sturctural 
notions. Thus, upon.in We rode/sailed/rushed upon the 
enemy incorporates the notion of 'attack', seemingly 
equivalent to the paraphrase 'into attack upon' The 
closed-class adverb tomorrow is equivalent to the 
phrase 'during the day that occurs next after the day 
during which I am now speaking', an example of an 
organized interlacing. 
7. One can note, for example, the effect on one's 
internal cognitive representation in considering first 
the sentence I looked at the dog and then I looked at 
the dogs. The addition of the grammatical element -s 
has a major effect on the delineational breakup of-- 
tp ~ut it vis~ally--the scene before the mind's eye. 
8. For example, augmentative and diminutive inflec- 
tions, insofar as they refer to actual size, seem to 
specify size relatively greater or lesser than the 
norm for an object. And grammatical elements spec- 
ifying distance (like English way and just appearing, 
e.g., before up there) appear to specify notions of 
'far' and 'near' that are relative to the current ': 
situation. 
9. It is true that there are the traditional terms 
"semelfactive" and "iterative" referring, respectively, 
to one and more than one instantiation of an event. But 
there is no real equivalent tonumber: "aspect" in- 
cludes too much else about the temporal structure of 
action. And in any case, none of the traditional 
terms refer generally to both the dimensions. 
10. The mechanism actually resorted to by both English 
and French in many such cases, including that of tear, 
is the use of the plural, as in: 
(i) TearLflowed through that channel in Hades. 
There seems t6 be a sequence of cognitiye oper- 
ations here in getting from a bounded to an unbounded 
quantity. Speculatively, the bounded quantity is 
first treated as a uniplex entity, it is then multi- 
plexed, the resultant entities are conceived as spa- 
tially juxtaposed, and their boundaries are lastly 
effaced. 
11. The present category may be prone to confusion 
with the preceding one. Contributory here is the 
normal meaning range of continuous, which as easily 
covers 'boundlessness' as it does 'internal seamless- 
ness' However, the two categories can vary indepen- 
dently. Thus, in the preceding section, the lexical 
examples given for unboundedness, water and sleep, hap- 
pened also to be internally continuous; but the same 
demonstration of unboundedness could have been made 
with internally discrete examples like timber and breathe 
12. Theredo exist certain mechanisms for such reversal. 
Thus, taking an unbounded case, the continuity-spec- 
ifying word water can be shifted toward being cognized 
as discrete by the locution particles of water, as in: 
(i) Water/Particles of water filled the vessel. 
However, the grammatical complex used here does not 
directly specify the shift but, like the one in Note I0, 
seems to involve a several-atage route of cognitive 
operations. 
13. For schematizing action along the one-dimensional 
time axis, an adaptation of the two-dimensional A, B, 
A, and B diagrams would be necessary--and can be 
readily visual~zed. 
14. The lexical types for several of these intersec- 
tions, it should be noted, do have traditional terms. 
Thus, nominal forms of the a, A, and B types, respec- 
tively, have been called count nouns, collective nouns, 
and mass nouns. And verbal forms of the a and B types, 
respectively, have been called punctual and durative 
verbs. The matrix presented here augments, systemat- 
izes, and generalizes the traditional notions. 
15. It may be considered an extension of the cate- 
gory of state-of-boundedness via the incorporation 
of the notion of uniplexity. 
16. This category might be considered an extension 
Or generalization of the "disposition of a quantity". 
Clearly, this category and the preceding five all belong 
together in treating the greater disposition 6f a 
quantity, but the relationships have not yet all been 
worked out. 
17. Our main purpose here is to note the shift in 
structure type. The shift in content, which will 
doubtless prove to have some regula~tv is not clearly 
understood at this point. 
18. A major function of perfect forms in language in- 
deed appears to be the one involved here. More par- 
ticularly, the perfect seems able to specify the temp 
oral counterpart of matter located within a bounded 
extent of space, as in (i). That is, a sentence con- 
taining the perfect, as in (ii), suggests a paraphrase 
like that in (iii), which is homologous with (i): 
(i) There were 5 aspirins on the table. 
(ii) I have taken 5 aspirins in the last hour. 
(iii) There were 5 aspirin-takings in the last hour. 
(In support of this interpretation, as pointed~.out to 
me by Peyton Todd, the perfect can be noted always to 
involve a temporal span bounded at both ends.) 
19. All three notion--identity of a quantity, portion- 
taking of a quantity, configuration of the portion-- 
are generally specified simultaneously.(or, "conflatedly" 
--see Talmy (1975)) by lexical items that would fit 
in the A/B row of (20). For example, (a) tear spec- 
ifies not only a certain shape of Quantum, but also the 
23 
material involved: lachrymal fluid. Such words gener- 
ally do not participate in an "NP of NP" construction 
--like *a tear of milk--unless they in fact accede to 
a shift toward the type of word represented in drop. 
20. There is a foursome of apt terms that can be ap- 
plied to the two levels of synthesis in the two direc- 
tions of shift, as indicated in (i). Employed here 
is the term "Figure" as it is used in my other work 
(Talmy 1978, 1976): 
(i) cluster: "composite Figure" 
trees: "multiple Figures" 
iceberg: "meta- 
Figure" 
2 halves: "component 
Figures" 
21. For the plural form oysters, the plural form si- 
phons is ambiguous as to whether there are one or more 
siphons per oyster. All the other combinations unam- 
biguously indicate the number of siphons per oyster. 
Thus, the exemplar form is always unambiguous in this 
reagard--one of its advantages over the full-complement 
form. This same arrangement holds through the list. 
22. I have long wondered what the differences between 
each and ev__ve~Tmight be. One apparent difference shows 
u-phere. Each seems to be the exemplar counterpart 
of all the but not of all without the (*Each oyster. 
has a siphon makes a poor generic assertion). 
Ts not constrained in this way, though it does strike 
me as more comfortably the counterpar t of all without 
the. 
23. One more pair can be added to this list by adjoin- 
ing two complementary unpaired forms from two different 
languages. The English form some, as in some friends 
of mine, requires the plural and has no singular coun- 
terpart. The Italian form ~ualque, as in qualque amico 
mio, requires the singular and lacks a plural. 
24. It seems likely that the language-related portions 
of the brain could have evolved to their present func- 
tions only in the presence of these already existing 
cognitive mechanisms and have incorporated their oper- 
ation. 
5. References 
Fillmore, C. 1975. An alternative to checklist theories 
of meaning. In: Berkeley studies in syntax and sem- 
antics, vol 1. University of California, Berkeley. 
..... . 1976. The need for a frame semantics within 
linguistics. In: Statistical methods in linguistics. 
Stockholm: Skriptor. 
Talmy, L. 1975. Semantics and syntax of motion. In: 
Syntax and semantics, vol. 4. J. Kimball, ed. Acad- 
emic Press. 
..... . 1976. Semantic causative types. In: Syntax 
and semantics, vol. 6. M. Shibatani, ed. Academic 
Press. 
.... . 1977. Rubber-sheet cognition in language. In: 
Papers from the 13th regional meeting, Chicago 
linguistic society. University of Chicago. 
..... . 1978. Figure and Ground in complex sentences. 
In: Universals of human language. Greenberg, Ferguson, 
Moravcsik, eds. Stanford University. 
24 
