ON THE INTERMEDIATE READINGS OF PLURAL NOUN PHRASES

ON

THE

ＩＮＴＥＲＭＥＤＩＡＴＥ△ＲＥＡＤＩＮＧＳ‥ＯＦ

犬 ニニPLURAL

ＮＯＵＮ＼ＰＨＲＡＳＥＳ……… ……

 二    十       Tsutomu Kato犬 ノ  ニ      ∧   く

（￡）ｅｐａｒｔｍｅｎt ofＥｕｒｏｐｅａｎ ａｎｄ Ａｍｅｒｉｃａｎ Ｃｕlｔｕｒｅ,Ｓｃｈｏｏｌ of刀umanitiesﾚａｎｄ Ｅｃｏｎｏｍｉｃｓ）

According to GiUon (1987√1992), (1) can have an intermediate reading, which is different from a collective･ reading or a distributive readingレ……… =･   十尚     ＼

  (1) Thesをmen wrote operasレ ．  ∧ ∧  ‥‥‥‥∧  ニ  ……：    ＝ ﾀﾞ‥‥ ‥‥‥

   Gillon argues that (1) willダbeダtrue when "these ｍ面"くdenotes the▽fo面individuals, Mozart, Handel, Gilbert and Sullivan. This reading cannot be collective because the four individuals did not write any opera in むoUaborati:on, and it cannot be distributive because neither Gilbert nOｒダSullivan wrote an opera on ｈｉsしｏｗｎレ耳ｅ裕,レｗｅ十needしtoexamine犬the relation between the denotation of "these men" and itｙﾄdivision:intoく面ｒ愧レ(2b) is the division of the set denoted byﾄ≒hese men'≒nto parts尚which makes (1) true.・・. ･.･つつｊ

(2) _a.

b.

{Mozart, Handel, Gilbert, Sullivan} {{Mozart}, {Handel卜∧{GilbertレSullivan}}

(2b) is ａ partition of (2a)･.1･(3a) and (3b) arりthe greatestﾚpartition･of (2a):and the･least

partition of (2 a), resうectively,皿dうtコis obvious that (3a)

corresponds to the collective

reading of (1) and (3b)･corresponds

to the･ distrib緋ive reading〉of(!)，∧   犬 犬

(3) a. {{Mozart, Handel, Gilbert, Sullivan}ト   b.上{{Mozart}, {Handel}, {Gilbert}, {Sullivan}}

Gillon,･ therefore, ･むoncludes that a plural noun phrase can have at lea昨 as many readingり as there are possible partitions of the set of individuals de叩加d by∧that noun∧phras･e. Hence, the number :of possible readings of a 卵ntence co皿aining aしplural noun phrase as its subject is determined based on the possible partitions of that pluraト subject noun phrase｡    Gillon, however, maintains六体at partition readings are not all the possible reading of a subject ･plura卜noun phrase. Consider theイollowing sentence.十一   ‥‥‥‥‥‥‥‥ ‥‥‥

   (4) These men wrote musicals.      ＞       十  尚

According to Gillon's interpretationトof this Seれtence, it is true賀hen "these men” denotes 仙ｅ three individuals, Rodgers,･Hammerstein ａｎｄ∧Hart.耳owever, anyﾀﾞpartition of the set of theseしthree individuals can make (4) true∧Rodgers and Hammerstein∧wrote m面icals in

208 _{Res. RepレKochiトUniv. Vol.43 (1994年) Hum}

collaboration, and Rodgers and Hart wrote musicals inごcollaboration. T畑se two groups, (Rodgers, Hammerstein) and (RodgersにHart)けhough√canno卜partition the set consisting of these three individualsレIn this case, Rodgers沁ａ com皿on member of theﾄtwo groups, and the intersection o卜the two sets representing these two groups is not犬empty. This kind o仁division ofよhe set, 塙町efore, cannot be per耳littedもy the definition of 皿rtition, w･hich requires that the intersection of any two sets be null. ＼Ｔ０.ｍａ･keし七histype of reading of (4) possible, Gillon maintainsトthat these tｗｏ犬groups minimally cover the Set十in question though they may not partition ｉソ In the end, GiUon (1992), which十is the developed version of Gillon (1987), assumes tha卜the numb町可possible readings ＼tｏ＼which a subject plural noun phrase is susceptible is the ｎｕホbeｒ上of possible minimaトcovers of the。set representing the denotation of that ｡subject plural･ noun phrase.   犬    ＼ し  ･｡    Gillon (1992: 617-619) introduces two terms√aggregate and aggregation, to simplify

the discussion about the range of possible readings of 雅 plural noun phrase. An: aggregate is defined as 乱n object formed from one or more members of ａ given background set.^ If

a background set has three distinct eleme叫s, ei√e2にり, then seven aggregates caれ∧be formed from these eleme皿s, ei, e2, es, eiez, eie^∠eie^ and 616263."トGillon, then, defines an 昭gregation ａｓ◇aset of aggregates whichﾙsatisfies the∧tｗ9ﾚconditions: the join of the aggregates in an･aggregation yields the greatest (or unit) a:面regate二飢ｄ ｎｏコ昭gregate in an aggregation is aへproper ･sub-aggregate of any other aggregate in the aggregationトThe aggregations formed from a background set ｡｛el√e2,ea}レtherefore, ￡ぼe{ei,ﾚj2ｿﾞe3｝･,{eie2.･ e^}, {eie3パ2｝しバeiez, ei}, {eie3, eie2}, {eie2, 6263}バeie3, e2り} and･｛岫海乱5=Ｇｍｏｎ（1992: 621) gives the followinぼ sentence and the possible大弓itｕ証ionsin which （5）ﾚcan be ･true.十，

(5) These men rowed.

I       ‥‥‥  ‥‥‥‥ : 十("Th･ese men" denotes Tom, Dick and Jerry.)

(6) Situations      ／   ･.  :･ ．･ 二 ＼

 ａ， Tom, Dick ･and Jerry were in one boat,ﾚｹﾞeach pulling an oar.………   b. Tom and Jerry were in one boat, at someしpoint･, each pulling皿oar; and Tom   ＼ and Ｄ沁k were in one boat, ａ卜some other time, each puUi昭an oar. ∧  十   cにTom and Dick were in one boat, each pulling anくoar; while Jerry was in  .. .皿other boat rowing.几   ………＼． ．･･  ．･． ･･  ．・． ・ ・． ． ･Ｊ 十   d. Tom was in one boat rowing; Di峰岸as in another boat rowing; Jerry was in  上still another boat丿owing･ ‥‥‥‥‥   ‥‥‥‥‥‥‥‥‥＝

The situations (6a)-(6d) correspond to the:aggregations of the denotation of "these men

づ7a)-(7d)√respectively.       ∧  ト      ト      し  ＼

ON THE INTERMEDIATE READINGS ＯＦ上PLURAL NOUN PHRASES (Kato)△209

ｄ。｛Ｔｏｍ,･ＤＩＣ瓦にＪｅｒｒｙ＼ _{(distributive reading)}

Ｗｈａ七is important here is that ｗｅしcan treatﾄtheしfour readings above,十especially the two intermed毎te readings (tha卜is, the minimal-cover reading andトthe partition寸eading卜quite

systematically in terms of aggregates and aggregatiと）ｎs√and the denotation of ａ subject plural noun phrase, which forms the background set 仔ｏｍ which aggrega七es and aggrega-tions are formed.        し         ノ   ト    ノ      ‥‥ ‥ ‥  犬Partition readings and minimal-cover才eadingsレplay ａｎ丿mport皿七万ole when we

interpret reciprocal sentencesよLeonard and･Goodman (1940) giveﾀﾞan interesting discussion of the logic of reciprocity concerning this point.         ト   十 ＼

Suppose we have as elements a set of three columns, each colored with three bands, as pictured in the accompanying diagram [＝(8)], andづSｕppo坤七hat the relation Sis S此h that "xSy"means that in some one band ―lower, middle or upper→the two entities ｘ ａｎｄ･ｙ are identically むolored (in ･ the sense that･ｎ=〇しＯ!or in tha卜band either in Xﾄor in ｙ 1S different from any color in that band in the other)ト IしiSC↓ear thatトS is ａ

relation like those already considered; that we may haveしthree columns五ke the ones pictured, such ａｓ ａＳｂ,ｂＳｃ,ａnd ａＳＣトeven though all three columns have no single color in any one band.･       上       ノ．  １   ．．．一一 ∧       ・･･ ･： ・

(8)

呼呼

The capital

川副ず﹂

呼呼万

 コ         ・distinct shades of color.･      .・ ･･.･･.･.   ･.

Langendφen (1978:190) makes the fo!lowing comment on this point.

If we cons七rue. the ･relation S to be like aｒｅｓimilaｒ t０，ａｓＬ面姐rd and Goodman's･ choicリニ）fschematic letter ａｎｄよsubsequent discussionしsＵぱest√we see that thむ situation diagrammed m (24)［= (8)］satisfies the∧ＴＣイTruthでondition) for the ERS (Elementary Reciprocal Sentence) (22)［＝（9几 but ｎｏレthaレof the corresponding ECRS (Elementary Covertly Reciprocal Sentence) （23）［＝（10）］し Inorder for the TC for the ECRS （23）［よ（10）］to be satisfied, a1トof the colors in some one ＼band would have to be identical.        犬 一一j‥‥‥  ‥‥‥‥‥‥‥

(9)

(10)

They are similar to one ａｎｏ七her. They are similar.    ト

210 _{Res. Rep. Koohi Univト Vol.43エ(1994年) Hum}

is true in its intermediate (more

precisely､minimal-cover)ﾚreading]becauseトａｎy十minimal-cover aggregation formed

from

the denotation of "they" ｍ･akes(9)∧廿臨斗＼ ‥‥‥ ‥

  Gillon

may

seemｸﾞto assert that any subject plural……nounphrase……is

susceptible to an

intermediate reading. Though

this 血町 be the△case.工臨ersohn(1989) argues against Gillon

(1987) giving the folio､wing sentences.し………=．．．＼六つ………

(11) a. The TAs were paid exactly ＼$7,000 1a鴎･year.△  づ b; The ･TAs were paid exactlyト$21,000 last year･･   c; The ＴＡs貰ere paid exactly･ ＄!4,000 las卜year.

According to Lasersohnﾄ(1989: 131), (11a) and (lib) are true, as Gillon's analysis predicts, 畑曲e situation where "John, Mary and Ｂｍ∧are the teaching即寸stants (TAs) in the local linguistics∧department" and ≒恥y ｗ6やら:paid $7,000 each last yeaピレIn thisダsituation, the denotation of "the TAが 陽ｎ恥ﾄrepresented asトthe setト{John√Ｍａｒ八万Ｂｍ}土庇d from this sｅt･we canニform two minimal coversﾕ,バ{John}√{Mary}√{Ｂ皿卜……al!･d･..･{･{John:,･･Mary,しBill}}ト Theイormer minimal coverコmakes (11a) true如causeしe収h……e!ement……of the cover was paid exactly ＄ 7･,000 last yeaぐａｎｄしtheしlatterｹﾞｍ江1血ａトcoveトmakesし(皿b)∧true because each element of the cover was paid exactly ＄ 21,0001卵卜yｅａ:ｒ， Lasersohn argues that what∧ls= important here is thatトGillon's analysis predicts that (lie)･is true in｡this s雨lation because

each element of the minimal cover {{John√Bill}, {Mary, Bill}} was paid exactly ＄14,000 last year. This reading, however,∧is impossib川畑this situation. Consequently Gillon's analysis is not correct･.上 / ＼    ………:･:     犬＼   ニ  ∧

･.・. ･･Lasersohn gives theイoUowing explanation to the point mentioned above. The possible readings of subject plural noun phrases are犬d6tりrmined by the verb phrases associated with tｈｅｍバｎthis case, ｏｎ!y the distributive reading or the collective reading is compatible ｗ社h the verb phrase denotation.プ耳ence, we cannoレget the minimaトcover reading in the situation mentioned above. Lasersohn (1989:133) gives the following explanation to the Rodge恥, Hammersteinしand Hart sentence. ………:    ダ         し   ト １

l am not･ sure l share Gillon's intuitions about 恰iS sentence, but ifｼ寸ｅ･wish to capture them :in the grammar, we can do so without major alterations to the theory through thQ･ use of lexicaトmeaning postulates･s牡ｐ皿面1賭i毎plications like that in (7)トwhere [砂庇d is the relation denoted bｙ=ｗｒitｅし(oni仙□collective reading), and w, x, y, andz ａｒeダgroups or individuals.     ‥‥‥‥‥‥ ‥‥‥‥‥‥‥‥‥  ‥   ‥‥

(7)：iwri￡ｄ(ω,y) & IωΓiteJCx,z) -*  ∧[writeKwロ工･ y u ･ｊ)

This meaning postulate ensures, for:instanceいthat if the groupごof Rodgers and Hammerstein wrote Oklahoma!and the group of Rodgers and Hart wrote Ｂａｂｅｓﾄ加 λΓ'ms, then the group of RodgersレHammersteinレ卵d∧耳ａれ wrote the group of

ON ＴＨＥ･

INTERMEDIATE

READINGS

OF PLURAL

NOUN

PHRASES

(Kato) ２!･１

…… However,▽we must be careful at this poi:ｎt∠What Gillonﾚmaintains is that one and the same sentence can have the reading which is neither distributive nor collective, and the possibility ｄ this reading can be･ascribed to theﾚ possible readinひof the subject plural ｎｏｕｎﾚphrase of the sentence. What Lasersohn maintains is that the situation mentioned

above can be described ･distributively ･or collectively but it･canno卜be described by any sentence which requires the mini血al-cover readingニwhich 1S neither distributive nor collective. In addition to this, Lasersohn's explanatioれto the Rodgers, Hammerstein, and Hart sentence is not so convincing as he suggests. When John and Mary met Tom, and John and Bill ｍ吋 Jane, can we say the group o卜John, Mary and Bill met theニ詐oup of Tom and Jane?尚 As for meanigﾀﾞpostulates, they should be abaれdoned if we大皿n find some principled ｗ孤ｙto eχplain the phenomena explained by them. ･.･..･･..･ ・.   ･･.     ･･.   Returning to Gillon's analysis, another type of counterexample to it is∧the sentence

which an anonymous refree of Lasesohn's･squib points to. (12) is仙e sentence at issue, and Lasersohn (1989:132) states that this sentence seems true in the situation where "John is the TA for Class A,トMary 1S the TA for Class十B, and John and May are jointly

appointed as the TA for Class C， and the standard pay for ａﾄＴＡい1S犬$7･,000."

（12）Ｔｈｅ TAs get paid exactly ＄ 7,000.

He also states∧that "... no minimal cover of the class of TAs seems appropriate,”and  "{{John}, {Mary}, {John, Mary}} is ｎ４ ａ minimal coverﾊﾞ' According to the definition of minimal･cover, this is unques:tionably true. ^ The point｡/4t issue is very obｖ･iou・づln･･one respectダJohn is ａ TA, and Mary is a TA. In another respect√the collectionニor aggregate of John and Mary is treated as ａ single TA. From another point of view, "the ＴＡ” denotes one individual in some cases, and it denotes tｗｏうndividuals in other cases. No definite noun phrase containing ａ singular common noun showsﾔsuch ａ十change in its denotation. The above situation which makes (12) true is quite exceptional in this regard if ”TA" is consideredﾚto be ａ common ｎｏＵｎ.＼If we regard”ＴＡ’「as a kind: of collective noun, this situation may be !essしexceptional. For instance↓ "the committe” denotes ａ collection of individuals and we cannot say how many individuals com皿沁ｅ the committe without referring to our extralinguitic knowledge,▽017 tWbトdifferent尚committees may consist of the same individuals. If we admit this possibility, the denotation ofドthe ＴＡs” may be less exceptional though we need to admit that ａ collectiv自大noun can have a singleton set as its denotation. In this むase, (12) can be true on the distributive reading of:"the TAs” in the situation above mentioned, and this reading is compaΓable t6 the distributive reading of "The families received exactly ＄ 7,000.”  ‥‥‥‥  ‥ ‥‥万     ＼    ∧…………    ・Gillon (1990) argリes against Lasersohn on this pointﾚ＼ＡＳ Gillon states, Lasersohn

maintains tｈt∧(lie) has only two readings: in one interpretation∧thむTAs collectively received $14,000, and i=ｎthe other interpretatioねthe TAs received individﾘally $14,000. Gillon (1990: 483), however, maintains that /”with 痛己 proper stage setting, the intermedi-ate readings l claim exist can be brought out into the openトif･･not for ＼the ｡.ｅχactsentence

212

Res. ＲｅｐレKochi Univ√Vol.43エ(1994年) Hum

in (16)[= (llc)], then at least for one suffiφentlyトlike上itto suぽest ｡any unavailability of 仙e readings for the〉sentence (16)寸十(lie)]十φ卵し畑十due如the adverbﾌﾞexactly∵'………

(13) The TAs were paid their ＄!4,000 last year.

T.hりfollowing situation is given to justify the claim.

A chemisty department has tｗ〇一七eachingassistants for each of its courses:,ﾚone for 七he recitation section and one for the lab section. TheトdeμΓt血ent h郎血ore than two teaching- assitants and it has set aside ＄ 昧000 f健二ｅ邸h course∧with teaching assistants. The total amount of ｍ叫町disbursed for them, then, is greater than

$14,000. At the same time, since the workload for teachins' a course's sectio［C叫1 vary from ｏ･nesection to anoﾌther√the department per血its each team of assistants for a course∧to decide for itself how to divide the ＄ 14,000 the team is to recieve. Suppose that it turns out, as it very well could∧under such circumstances, 轍ａt◇皿teaching asssitant is paid exactly ＄ 14,000. Yet, it seems to∧me犬thaレeither of the sentences in (16)く=aic)]or (17)[〒(13)]could beﾄtruly affir血ed, though neither sentence, by hypothesis,･is true in virtue ｡0f either a collective･or a distributive reading.

   The following points are important in this supposed situation t6 get an intermediate reading though it is riot c!ear ih〕his situation whether that intermediate reading １Ｓａ partition reading or ａ minimal-cover reading.し    ‥‥‥‥‥‥‥‥‥‥‥＼

   （Ａ）Ｔｈｅtotal amount of money disbursed for the TAs is more thaれ$14,000.￢ヤ（13）     cannot have a collective reading becaﾘSeトthe total amount paid must be explicitly   し expressed to get a collective reading. ＼ ⊃    ‥‥‥‥‥

   （Ｂ）ＮＯTA received S 14,000.→(13) canno卜have a distributive reading because ･曲e ＼    amount paid to an individual must be explicitly expressed to get a distributive   し  reading.      /     犬       /        ヶ

  (C) $14,000 is set aside for ａ course with two 邸Siりtants, and∇two assistants comprise    a team.→Ａ team consisting ･of two TAs is paid $14,000.  ‥‥‥‥‥I ＼

ＴｈｅＴｍｏＳレimportant point seems to be (C) because this◇ｃ叫d雨on determines the correspondence between $14,000 and a team consisting〉of two TAs. The logic used here, howeverパs very similar to the logic which I touchedレon above√トWhat 1S reallyエmeant by （!3) in this supposed situation is 曲ａt each of∧the teams consisting of two ＴＡＳ:･received

exactly ＄ 14,000. This reading is, of course, distributive.犬The point, however√is that "tea ｢’iS a collective noun. Henceパf we know that:a team consists of tｗｏｹTAs, we may recognize thatフthe TAS” has the same extension as "the tea ｢≒i.Qレ七恥犬set of theべTAs. In this case ＄ 14,000 iS＼associated with a team ･(= two ＴＡＯ丿血dフtheir”ｍａトhelp the

establishment of this association through its anaphoric nature 七こ）塙e subject.: "Exactly” seems a kind (ﾋ）fabsolute standard against which amount can beくmeasured. The candidate for that standard seems to be an individual ０れ a set, and not a partition or a minimal cover of ａ se･      ･.       一犬・ ■ ■       ■■

ON

THE

INTERMEDIATE

READINGS

OF PLURAL

NOUN

PHRASES

(Kato) ２１３

   The above discussion shows the difficulty in大如tting intermediate readings of plural noun phrases and the一一sentences containing themレlf a pluraトnoun phrase has∧a set of individuals as its denotation, and the members are structured into aggregates formed from

them, the most salient elements are supposed to be minimal aggregatesﾆand the unit aggregateﾝMinimal aggregates are salient because they consist of a single individリal, and indi･viduals面己ontologically･basic units. トThe unit aggregate is salient :because｡汀｡the n･oun phrase which gives the background 印t 1S コdefinite, it denotes the maximal set inニthe relevant context. This means that they are eS卵ntially salientレand nothingﾆIS血ecessary〉to make them salient.トIt ｆｏｎりws then that the two readingsレdistributive 四d collective, are more easily available. To' get intermediate readings we have to make the aggregates corresponding to these readings salient because these aggregates have no constituency principle and th町 are only collections of the elements which∧happen toﾚbe treated as one objectレNaming may be a possibleトmethod to make these aggregates salient as the◇above discussion about･ collective nouns shows. Collaboration of the:individuals denoted by the ａぽgregates may be ano七her factor, and we can name the individuals which correspond to these aggregates, e.ｇ. ｃｏａｕtｈｏｒｓレSometimes predicates require intermediate readings of plural noun phrasesにe.g. reciprocal sentences.'" These conditions on the intermediate readings ｒ:equire complicated situationsニand this leads to the difficulty∧in obtaining these ｒeadingS.＼.. ･･   ・. ・･..・・ ・・・        ..  ・・   .・・  ..・  .･.  ･･.  ..･.・ ･･レ

 犬  ｊ      NOTES     し        ニ●● ● ●●●●●

 ニ･１ Gillon (1992:j 616, footnote 14) gives the foUowi昭definition ofトpartition. A partition is a family of sets, each of which is a non-empty subset of a given sｅt√distinct sets in which family are disjoint and the union of which is the given set. This （斑n be put

more formally as follows:      ＼       ＼ ：  犬      ‥‥‥‥‥ ‥  ‥‥

(i) X partitions Y iff  十      Ｉ   ＼

 χ⊆Ｐ（Ｙ）∧口ダＸ∧ＵＸ午Ｙﾉ＼ｙｘ,ｙ亡Ｘ（ｘｎy上≠μ→X

=

y)

 (where

"P(Y)"

means

”thepower set of Ｙり＞  :＼   ＼

  2 Gillon (1992: 617, footねote 15) gives･曲ｅ･ｻfollowing definition.八:covむr is just山ke a partition eχcept it is not restricted tｏしdisjoint如体  ＼ ＞ ・.・  ･･ ・ I.  ・  ･･・

  (i)χ covers Y iff χ⊆Ｐ(Ｙ)一八･口∈三十χΛＵχ＝Ｙ｡．・･･．． ・･．       ･．

A minimal cover of ａニset is ａ smallest family ofしn6かempty subsets of =ａset which still manage to むover it. ト  ト         十 ＼       ト ‥ 才六   ニ

(ii) X minimally

covers Y iff ｘ covers･Υ Λ… ……

＼(∀名)べ(Z

covers Ｙ 八X

covers Z)→Z.＝IX)･

214

Res. Rep. Kochi Univ. Vol.43 (1994年）Ｈｕｍ

   3 This object may be abstractニor りoncreteにdepending o!1 the property of a given background Se札 ．． ……   ………ｌ･ ﾉ‥‥‥‥‥‥  ‥‥‥‥‥    ‥‥‥

   ４プItis obvious

that the set consisting of these aggregates has the complete join

semi-lattice structure with a unit 卵d

without

zero………This

is∧diagrammically represented

belowレ      犬  ヶ ＼    ･●     ‥‥‥ ‥ ‥  ‥‥

Background

set 十{ei, ez, 63}

uｎ八丁(ｏｒgreatest)ﾚ aggregate

eie-i     ・ eies  一一  e2e3･

     ／   犬ei      e2      e3    minimalaggregates

Many

scholars adopt this kind of lattice structreをs ａ struc･turewhich柿e

set denoted: by

a plural noun

phrase has. For instance, Barkerべ1992:

76) giv:eｙ柿色following structure

formed

from

the denotation of "theしｍｅｎ”

(=

"JohnレBill, and To

｢’）.上   尚 ＼

Proper

sums

Atoms

ｊ十ｂ十ﾆt

5 The following diagram

represents t恥ｿrelations among

these aggregations

{eie3, eie2}    {eie2, e2e3}   ･{eiea, ezei]｡

{eie2, es}十     仏e3｡岫 ニ  {e2e3, ei}

{eie2e3}

We can easily find that when七he denotation of a pluね1 nounﾚphrase is the set {ei, ez, e3}, its collective reading corresponds to the aggregation ＼{e1尚雌}レand its distributive

reading-ON THE INTERMEDIATE READINGS OF PLURAL NOUN PHRASES （ＫＡＴｏ） ２１５

corresponds lto the aggregation {ei, e2, e3}.･The ”･partition” readiねgS･ correspond ･ tｏ･･the aggregations血the second row (fro血bottom祐top), and theダ"minimal-cover'「readings correspond to the aggregations 毎 the third rowト       ＜     ＼   ニ  ∧，

 〉 6 Cf. Note 5. This case may seem to be very special. However, if七here is something special in this case, I believe that it can be ascribed to the fact･ that （9レis ａ reciprocal sentence. In any case, some kind of minimal-cover犬reading can satisfy the truth cond姐on

for (9). Langendoen (1978) also discusses the relation between 柿e partition readings of subject plural noun phrasesコand the truth conditions for reciprocal sentences.      ∧

   7 Itis obvious that the clear definition of the relation between an individual and the

singleton selレconsistingof that ･individual is neces･sary.       一一 ……

  1 8 This point will be discussed below.   ∧     ‥‥‥:  ．．       十

  9 Cf. Note 2／       ト        、

  10しMini耳lal-cover readings seem to be more difficult to obtain t姐ｎ partition readings. This may be so because there are expressions which require partition readings of

a plural noun phrase e･g. tｕ)ｏ b:ｙ tｕ)ｏ、b^y ｔｌｉｅ （ｋ２ｅ瓦、but to my knowledぽe、 there is no expression which requires minimal-cover readings.   ∧       ニ

REFERENCES

Barker, Chris (1992) "Gr･oup Terms in English: Representing Groups as Atomぐ' ｍ Ｊｏｕｒｎａｌ ｏｆ   Ｓｅｍａｎticｓ ９， 69-93.

Gillon, Brendan S. (1987)ブThe Readings

/of Plural Ｎｏｕｎ｡ Phrases,”＼ｍ Ｌｉｎｇｕiｓticｓ:Ｃし几ｄ    Ｐｈ.iloｓｏｐｈ.'ｖ10,199-219.       ∧       十

(1990) ."Plural Noun

Phrases

and

Their Readings: A

Reply

tｏ〉Lasersohn,”in

（!992) "Towards

ａ Common

Semantics

for English Count

and Ｍ卵s Nouns,”in

   ￡ingiiis￡ics and Philosoi］みｙ 15, 597-639.        ･..・･. ･･ ・.y ･･. .・・

Kato, Tsutomu (1991) "A Note on the Non-collective｡Non-distributive Readings of Subject    Plural Noun Phrases,”inＲｅｓｅａｒｃｈ ＲｅｐｏｒtｓｏｆＫｏｃｈ-ｉＵｎｉＤｅｒｓiｔ^? 40，Ｈｕｍｏしnitieｓト174-182.

(1993)”On the Readings of English Plural Noun Phrases," inＲｅｓｅａｒch, Ｒｅｐｏｒｔｓｏf     Ｋｏｃｈ,iUniｕｅｒｓiり42，Ｈｕｍａｎｉｔieｓ,221-228.         ∧       ＼ し ‥ ＼

Langendoen, D. Terence (1978) "The Logic of Reciprocity,” inＬｉｎｇｕiｓticIn.ｑｕiｒ'ｖ 9,177-197レ LasersぐDhn, Peter (1989) "On the Readings of PluraトNoun Phrases,”in Ｌｉｇｕiｓｔic･Inｑｕiｒｙ･20，     130-134.し  ＼      十      イ

216

Res. Rep. / Kochi Univ√Vol.43∧(1994年)･Hum

Leonard, Henry S√and Nelson Goodman (1940)プThe Calculus∇of Individuals and Its:Uses,?’in   しＴｈｅ ＪｏｕＴＴｉａｌ可Ｓツ男boｌｉｅＬｏｇｉｃ５， 45-55…… …………万………:‥‥ ‥‥‥‥‥‥‥‥

Manuscript received: September 30， 1994 .. ･..・  ・Published:･December 26,･1994