Pima Quantifiers

Abstract

After presenting some basic genetic, historical and typological information about Pima this chapter outlines the quantification patterns it expresses. It illustrates various semantic types of quantifiers, such as generalized existential, generalized universal, proportional, definite and partitive which are defined in the Quantifier Questionnaire in Chapter 1. It partitions the expression of the semantic types into morpho-syntactic classes: Adverbial type quantifiers and Nominal (or Determiner) type quantifiers. For the various semantic and morpho-syntactic types of quantifiers it also distinguishes syntactically simple and syntactically complex quantifiers, as well as issues of distributivity and scope interaction, classifiers and measure expressions, and existential constructions. The chapter describes structural properties of determiners and quantified noun phrases in Pima, both in terms of internal structure (morphological or syntactic) and distribution.

13.1 Introduction

This chapter discusses quantification in Pima, a Uto-Aztecan language spoken in central and southern Arizona.¹ It is mutually intelligible with the better studied dialect Tohono ‘O’odham (Zepeda, 1983). Few studies of quantification in this language have been undertaken, the most readily available being the description of quantifier float in Munro (1984).

Before describing the quantifier patterns, some basic familiarity with the language is necessary. Pima is a quintessential ‘non-configurational’ language. Indeed, its sister dialect Tohono ‘O’odham was one of the original languages used by Hale (1982) and Jelinek (1984) to argue for this class of languages. I point this out, not to make a claim about the proper theoretical analysis of the data to come, but to give some typological expectation of the patterns to be encountered. All six logical permutations of subject, object, and verb are possible, with interpretive differences (if any) lying largely in the information structure (Hale, 1992; Payne, 1992). The sentences in (1) are adaptations of those given by Hale (1992) for a different dialect. Flexibility of word order extends into the major constituents, so both possessor-possessum and possessum-possessor orders occur (2, 3). Also, both prepositional and postpositional structures are possible for the same adposition (8, 9). (In some cases, adpositions split the object phrase, creating a kind of ‘impositional’ structure.) There appears to be little to no effect on relative scope or binding possibilities based on different orderings of the major syntactic constituents. This flexibility among the lexical constituents is not reflected as strongly in the functional constituents, where the ordering of elements is more strict. The most consistent word order pattern is the presence of a second position auxiliary, which encodes subject agreement, aspect, and modality. Only constituents can appear in pre-auxiliary position. The second position pattern can be seen in each permutation in (1).²

(1)	a.	Vakial	’o	heg	vipsilo	ha-cecposid (SOV)
		cowboy	aux	det	p,calf	3p-p,brand
		‘The cowboy is branding the calves.’
	b.	Vipsilo ’o ha-cecposid heg vakial. (OVS)
	c.	Ha-cecposid ’o heg vakial heg vipsilo. (VSO)
	d.	Vipsilo ’o heg vakial ha-cecposid. (OSV)
	e.	Ha-cecposid ’o heg vipsilo heg vakial. (VOS)
	f.	Vakial ’o ha-cecposide heg vipsilo. (SVO)

(2)	a.	heg	John	kalit
		det	John	car
		‘John’s car’
	b.	heg	kalit-aj	heg	John
		det	car-3poss	det	John
		‘John’s car’

Null anaphora is pervasive: independent pronouns are optional as arguments of a verb, possessors, and objects of adpositions.³ It is not uncommon for a sentence in a narrative to lack any nouns whatsoever. Person and number (to a lesser extent) are usually recoverable via agreement morphemes found on the auxiliary and verb (compare 3 against 1), possessum (4 against 2), and adposition (5).

(3)	ha-cecposid	’a-ñ
	3p-p,brand	aux-1 s
	‘I am branding them.’

(4)	heg	kalit-aj
	det	car-3poss
	‘his/her car’

(5)	heñ-wui
	1 s-to
	‘to me’

13.2 Quick Overview of Quantifier Patterns

13.2.1 Overview of D-Quantifiers

The d-quantifiers in Pima are not determiners; they are adnominal expressions that may occur within the determiner phrase. The form of the DP is strongly influenced by where in the larger syntactic structure the phrase appears. There are four elements that distribute as determiners: the demonstratives ’iid a ‘this’ and hega’ i ‘that’, a specific indefinite (with some unclear semantic issues) ge, and a ‘default’ determiner heg. While the first three determiners appear according to the meaning, the presence or absence of heg appears to be mostly determined by syntactic position.⁴ Heg is used when the DP is not in certain syntactic configurations, including sentence initial, before a selecting adposition, before a selecting possessum, and when serving as a main or secondary predicate. In most other cases, heg is required to be present. There is no apparent change to the meaning regardless of whether or not heg is present (Hale, Jeanne, and Platero, 1977; Fitzgerald, 1994).

(6)	Keli	’a-t	’am	hii
	man	aux-pf	fr	see(pf)
	‘The man went there.’

(7)	M-a-t	hii	heg	keli
	fr-aux-pf	see(pf)	det	man
	‘The man went there.’

(8)	Kii	’amjed:
	house	from
	‘from the house’

(9)	’amjed	heg	kii
	from	det	house
	‘from the house’

Definite, indefinite, and generic DPs are not morphosyntactically well-differentiated. Generics must be plural, but that appears to be the sole restriction placed on one of the three but not the others (10, 11). The addition of demonstratives or quantifiers can make definiteness or indefiniteness more explicit, but there are no words or patterns to explicitly mark generics. In all three types of DP, the determiner heg can used if the word order allows it.

(10)	Gogogs	’o	tototk
	p,dog	aux	p,bark
	‘The dogs are barking.’, ‘Some dogs are barking.’, ‘Dogs bark.’

(11)	Gogs	’o	totk
	dog	aux	bark
	‘The dog is barking.’, ‘A dog is barking.’, *‘Dogs bark.’

This means that in negative sentences, a simple DP can be definite and scope out, or indefinite and (possibly) scope under negation.

(12)	Pi	’a-ñ	ha-ñeid	heg	‘u’’uhig
	not	aux-1 s	3p-see	det	p,bird
	‘I don’t see the birds.’, ‘I don’t see any birds.’

D-quantifiers are usually added into the DP before the noun and after a determiner, if any (10). This changes in partitives, which will be discussed below (Section 13.6.4). However, the language prefers to float quantifiers whenever possible (13, 14). The lack of heg in a wide range of cases and the frequency of floating means that it is rare to see a d-quantifier clearly in the middle of a determiner phrase.

(13)	Suzanne’a-t’am’iha-gi’igheg gook ‘i’iks
	Suzanneaux-pffrinc3p-shake(pf)dettwop,blanket
	‘Suzanne shook (the) two blankets.’

(14)	Suzanne’a-t’amgook’iha-gi’igheg ‘i’iks
	Suzanneaux-pffrtwoinc3p-shake(pf)detp,blanket
	‘Suzanne shook (the) two blankets.’

13.2.2 Overview of A-Quantifiers

A-quantifiers function as adverbs, generally appearing before the verb, but with significant flexibility. (15) shows a typical pattern. (16) shows an adverb fronted before the auxiliary.

(15)	Gogogs	’o	gokko	tototk
	p,dog	aux	twice	p,bark
	‘The dogs barked twice.’

(16)	Shel	’a-ñ	hem-veehejed:	hihidod:
	always	aux-1 s	2 s-for	cook
	‘I always cook for you.’

13.3 Existential Quantifiers

13.3.1 Existential D-Quantifiers

Momentarily setting aside the indefinite pronouns, there are four words with existential semantics: hema ‘one, a, some (singular)’, ha’ i ‘some (plural)’, mu’ i ‘many’, and the specific indefinite ge. The first three distribute like standard d-quantifiers, the last more like a determiner. For the examples below, recall that the default determiner heg is missing from sentence initial contexts.

(17)	Hema	gogs	’o	totk
	a	dog	aux	bark
	‘A dog is barking.’

(18)	Ha’i	gogogs	’o	tototk
	some(p)	p,dog	aux	p,bark
	‘Some dogs are barking.’

(19)	Mu’i	gogogs	’o	tototk
	many	p,dog	aux	p,bark
	‘Many dogs are barking.’

The most frequently encountered context for ge is to introduce unique and significant individuals into a narrative.

(20)	Gam-husha’ina’am-a-shge ce’ul’amohebiikeek
	gfr-farintnsextentc-aux-hrsycertainwillowfrirrwhere stand
	‘Long ago, where there stood a willow tree,...’

13.3.1.1 Cardinal Quantifiers

The native monomorphemic cardinal quantifiers cover the numbers from ‘one’ to ‘nine’. The words siant ‘hundred’ and miil ‘thousand’ are borrowed from Spanish. The word for ‘ten’, vest-maam, is derived from the phrase vees maam ‘all fingers’. Multiples of tens, hundreds, and thousands, are expressed using the frequentative form of a number (= frequency adverb, see Section 13.3.2), and the ones place is added using gami, a shortened form of the distal locative adverb gama’ i ‘over there’.

(21)	gokko	vest-maam	gami	gook
	twice	ten	over.there	two
	‘twenty-two’ (lit: ‘twice ten [and] two over there’)

13.3.1.2 Indefinite Pronouns

Counting interrogatives as a type of indefinite, the indefinite pronouns come in three parallel sets (Table 13.1). The exact syntactico-semantic distinctions between the sets are unclear at times, but there are some generalizations to be made. The two sets that occur in declarative sentences seems to correlate best with specificity, or identifiability. I therefore refer to them as the specific and non-specific indefinite pronouns. The non-specific indefinite pronouns can also be used as alternatives to the interrogative pronouns in constituent questions (Section 13.3.1.3).

Table 13.1 Some indefinite pronouns in Pima

Consider the following two cases. Both are existential questions, differing only in which set the indefinite pronoun is drawn from. There may be a greater suggestion that one could identify the individual in (22) but not (23); but this is not strictly necessary. (It is not at all clear whether the difference in the English translations reflects the same difference in the Pima.)

(22)	No	’am	hema	ha’icug	kii	c-’ed:?
	q-aux	fr	someone	exist	house	unposs-in
	‘Is there someone in the house?’

(23)	No	’am	hed:a’i	ha’icug	kii	c-’ed:?
	q-aux	fr	someone	exist	house	unposs-in
	‘Is there anybody in the house?’

The two sets are also used differently depending on the polarity of the sentence. In affirmative assertions, the specific set tends to be used (24), while in negative assertions the non-specific is used (25). Again, this is a tendency, not an absolute rule. The exact reasons for the choice between indefinite types are as yet unclear.

(24)	M-o	hema	ha’icug	kii	c-’ed:
	fr-aux	someone	exist	house	unposs-in
	‘There is someone in the house.’

(25)	Pi	’am-hu	hed:a’i	ha’icug	kii	c-’ed:
	not	fr-far	someone	exist	house	unposs-in
	‘There isn’t anybody in the house.’

13.3.1.3 Interrogatives

As mentioned above, interrogatives come in two sets: the wh-words and the non-specific indefinites. The choice of which to use appears to be largely syntactic. Wh-words are obligatorily fronted to sentence initial position. In such cases, the second position auxiliary usually encliticizes to the wh-word (26, 27).

(26)	Doo-p-t	naam?
	who-2 s-pf	meet
	‘Who did you meet?’

(27)	Baa-t	hii	heg	Rebecca?
	where-pf	go(pf)	det	Rebecca
	‘Where did Rebecca go?’

The non-specific pronouns are used when the constituent is not fronted. They may occur sentence initially, but they are not moved there by obligatory wh-movement. The most common occurrence of a non-specific pronoun as interrogative is when wh-movement is blocked. In one common pattern, the pre-auxiliary position is filled by a complementizer ku-, which serves to connect the sentence to the broader discourse. In such a case, there is no empty initial position to move a wh-pronoun into, so the indefinite is used (28). Similarly, in embedded questions the initial position is filled with a complementizer, so the indefinite appears in-situ (29). Movement is blocked when part of a conjunction (30). Echo questions (31) and multiple wh-questions (32) also appear to lack wh-movement, though these patterns are less well understood.⁵

(28)	Ku-s	hascu	ha-nolav	heg	Melissa
	c-dub	something	3p-buy(pf)	det	Melissa
	‘I wonder what Melissa bought.’

(29)	S-maac	’a-ñ	m-a-p	hed:a’i	ñeid	tako
	stat-know	aux-1 s	c-aux-2 s	someone	see	yesterday
	‘I know who you saw yesterday.’

(30)	Jason	c	hed:a’i	’am	heñ-ñeid?
	Jason	and	someone	fr	1 s-see
	‘Jason and who see me?’

(31)	Hed:a’i	ha-nolav	heg	komkjed:
	someone	3p-buy(pf)	det	turtle
	‘Who bought a turtle?’

(32)	Hascu	hed:a’i	ha-’ees?
	something	someone	3p-steal(pf)
	‘Who stole what?’

If there is a difference in meaning between questions with wh-pronouns and non-specific indefinite pronouns, it is one of how the questioned constituent relates to the broader discourse, not the semantics of the question form itself. The following pairs of examples are reported to be synonymous.⁶

(33)	Doo-t	o	mua	heg	kooji?
	who-pf	irr	kill	det	pig
	‘Who will kill the pig?’

(34)	Ku-t	hed:a’i	o	mua	heg	kooji?
	c-pf	someone	irr	kill	det	pig
	‘Who will kill the pig?’

(35)	Shacu-p-t	’aagc	o	mua	heg	kooji?
	what-2 s-pf	saying	irr	kill	det	pig
	‘Why are you going to kill the pig?’

(36)	Ku-p-t	hascu	’aag	o	mua	heg	kooji?
	c-2 s-pf	what	saying	irr	kill	det	pig
	‘Why are you going to kill the pig?’

13.3.2 Existential A-Quantifiers

There are three existential a-quantifiers that are not productively constructed: hekid ‘some time’, hebicuc ‘sometimes’, and hemho ‘once’. All others are derived. ‘Never’ is produced by negating hekid ‘some time’.

(37)	Pi	’a-ñ	hekid	hoohid	heg	John.
	not	aux-1 s	some.time	like	det	John
	‘I never liked John.’

Frequency adverbs are derived by suffixing -ko to a d-quantifier, e.g., gokko ‘twice’ (< gook ‘two’) and vaikko ‘thrice’ (< vaik ‘three’). This pattern is fully productive. Even syntactically complex numbers can take the -ko suffix, e.g., gokko vest-maam gami vaikko ‘twenty-three times’ (< gokko vest-maam gami vaik ‘twenty-three, lit. twice ten over there three’). Non-numerals also feed the pattern: mu’iko ‘many times’ (< mu’i ‘many’).

13.4 Universal Quantifiers

13.4.1 Universal D-Quantifiers

There is only a single universal d-quantifer, vees ‘all’. It can be used with either a singular or plural restriction. If used with a singular, it quantifies over the totality of the object (38); if used with a plural, it quantifies over members of the set (39). The number of the object in the following examples can be seen from the agreement morphology on the verb.

(38)	Vees	huu	’a-t	heg	pas-tiil.
	all	eat(pf)	aux-pf	det	pie
	‘He ate the whole pie.’

(39)	Vees	ha-huu	’a-t	heg	pas-tiil.
	all	3p-eat(pf)	aux-pf	det	pie
	‘He ate all the pies.’

The restriction can also be a coordinated expression, in which case the range covers all members of each conjunct. So, in the following example, every man, woman, and child vacated the city; nobody remained.

(40)	Vees kekel, ‘o’oki, c ‘a’al’a-tdaagtohegkiihim.
	allp,manp,womanandp,childaux-pfleavedettown
	‘All men, women, and children left the city.’

Non-specific indefinite pronouns can receive a universal interpretation when the head of a free relative clause. The relative clauses in the following examples begin with m-, the complementizing proclitic to the auxiliary. The non-specific indefinite pronouns are in the first position syntactically available, given that the auxiliary has to be the second constituent and the complementizer takes the first position.

(41)	’Am	g	cindat	m-a-p	hed:a’i	ñeid!
	fr	imper	kiss	c-aux-2 s	someone	see
	‘Kiss whoever you see!’

(42)	’Am	g	ha-nolav	m-a-p	hascu	’i-tatcua!
	fr	imper	3p-buy(pf)	c-aux-2 s	something	inc-want
	‘Buy whatever you want!’

13.4.2 Universal A-Quantifiers

Just as there is only a single universal d-quantifier, but universal interpretations can be assigned to other constructions, so it is also with the a-quantifiers. The single universal a-quantifier is shel ‘always’. Like the English translation, this word is frequently used hyperbolically, so that translating it as ‘most of the time’ or ‘often’ more accurately reflects the real usage.

(43)	Shel	’a-ñ	ha’icu	s-maac.
	always	aux-1 s	something	stat-know
	‘I always know something.’

(44)	Shel	’a-ñ	’absh	’am	’i-keishpa.
	always	aux-1 s	just	fr	inc-walk
	‘I always just walk.’

Other universals can be built from the indefinite hekid ‘some time’. This can either be done through prefixing with vees ‘all’ or cum ‘any’. Both phrases seem to have the same meaning as shel ‘always’, though it is likely subtler shades of meaning could come to light with further research.

(45)	Cum	hekid	’a-ñ	ha’icu	s-maac.
	any	some.time	aux-1 s	something	stat-know
	‘I always know something.’

(46)	Vees	hekid	’a-ñ	ha’icu	s-maac.
	all	some.time	aux-1 s	something	stat-know
	‘I always know something.’

Clauses can get a universal reading if the verb is in the habitual with no temporal adverbs. The first sentence below (47) shows a sentence with an implied specific time reference, the second (48) a more universal reading. This universal interpretation is an implicature, though, not an entailment, as can be seen in (49), where a temporal adverb cancels the implicature.

(47)	Heñ-hiksh	’a-ñ	m-a-n-t	’am	heñ-hihiviu.
	1 s-cut	aux-1 s	c-aux-1 s-pf	fr	1 s-shave(pf)
	‘I cut/was cutting myself when I shaved.’

(48)	Heñ-hikkash	’a-ñ	m-a-ñ	’am	heñ-hihivium.
	1 s-cut(hab)	aux-1 s	c-aux-1 s	fr	1 s-shave
	‘I (always) cut myself when I shave.’

(49)	Hebicuc’a-ñheñ-hikkashm-a-ñ’amheñ-hihivium.
	sometimesaux-1 s1 s-cut(hab)c-aux-1 sfr1 s-shave
	‘Sometimes I cut myself when I shave.’

13.5 Proportional Quantification

13.5.1 Proportional D-Quantifiers

Pima has no monomorphemic proportional d-quantifiers. It does, however, have an idiomatic expression ’ed:a hukka m ‘half’, literally ‘within the edge’. This expression distributes like a partitive d-quantifier (50), except that it has not been observed to float. (Grammaticality judgments on floating of ’ed:a hukka m are not available.) Note that if one uses a non-partitive construction (see Section 13.6.4 for details), ‘half ’ appears to modify the type of individual, rather than the quantity, even if that meaning does not make much sense (51).

(50)	’Ed:a	hukkam	heg	‘a’al	’a-t	’am	ñei.
	inside	edge	det	p,child	aux-pf	fr	sing(pf)
	‘Half of the children sang.’

(51)	#’Ed:a	hukkam	‘a’al	’a-t	’am	ñei.
	inside	edge	p,child	aux-pf	fr	sing(pf)
	‘The half-children sang.’

13.5.2 Proportional A-Quantifiers

There are no basic proportional a-quantifiers in the language. Proportional interpretations are generally created productively through modification of the universal and existential quantifiers. There is one idiomatic expression composed of a complementizer and particle, ku-...hiva, that together are interpreted as ‘usually’.

(52)	Ku-ñ	hiva	‘ii’e	heg	kavhii	sisalmad.
	c-1 s	prt	drink	det	coffee	p,morning
‘I usually drink coffee in the morning.’

(53)	Ku-p	hiva	memd:a	Cuk-shon	wui.
	c-2 s	prt	drive	Tuscon	to
‘You usually drive to Tuscon.’

13.6 Complex Quantifiers

Pima productively constructs complex quantificational phrases based around the core quantifiers.

13.6.1 Approximate Values

The most straightforward cases of complex quantifier constructions involve adding an adverb to slightly modify the value. Sha ‘about’ indicates that the value expressed next is an approximate figure (54). Cemalo ‘almost’ means that the quantity falls just short of expectation (55).

(54)	Sha	vees	‘a’al	’o	ñe’e.
	about	all	p,child	aux	sing
	‘Just about all children sing.’

(55)	Cemalo	vest-maam	‘a’al	’o	ñe’e.
	almost	ten	p,child	aux	sing
	‘Almost ten children are singing.’

These same modifiers can be applied to a-quantifiers.

(56)	Cemalo	shel	’a-ñ	memlicud	cikpan	wui.
	almost	always	aux-1 s	drive(hab)	work	to
	‘I almost always drive to work.’

(57)	Sha	hebicuc	’a-ñ	memlicud	cikpan	wui.
	about	sometimes	aux-1 s	drive(hab)	work	to
	‘I mostly drive to work.’

Imprecise quantities can be intensified, for example with si ‘very’ (58). The more emphatic shi ‘very (emphatic)’ often carries judgmental undertones (59). The judgmental tone can be emphasized by further adding ’abs h ‘just’ (60).

(58)	John	’a-t	si	mu’i	hemajkam	ha-wui	ñe’o.
	John	aux-pf	very	many	person	3p-to	speak(pf)
	‘John spoke to very many people.’

(59)	John’a-tshi mu’ihemajkamha-wuiñe’o.
	Johnaux-pfvery(emph)manyperson3p-tospeak(pf)
	‘John spoke to too many people.’

(60)	John’a-t’absh si mu’ihemajkamha-wuiñe’o.
	Johnaux-pfjustverymanyperson3p-tospeak(pf)
	‘John spoke to too many people.’

Cum ‘any’ precedes non-specific indefinite pronouns, serving as a domain widener.

(61)	Va-n-t	o	cum	hascu	ha-nolav.
	fut-1 s-pf	irr	any	something	3p-buy
	‘I’ll buy anything-at-all.’

(62)	Cum	hebai	daash!
	any	somewhere	put
	‘Put it anywhere!’

13.6.2 Comparative Quantities

Quantities can be compared using the standard comparative construction. The comparative meaning ‘more/greater than’ is formed with the adposition ba’i c ‘in front of’ and the proximate locative proclitic ’ i (63). ‘Less/fewer than’ is expressed with veeco ‘under’. Equality in the comparison is expressed with maasma ‘like, the same as’. In all three subtypes, the standard of comparison is introduced with mam ‘than’.

(63)	Eric	’o	ba’ic	’i-cuatk	mam	’aa ñ i’i.
	Eric	aux	in.front	pr-tall	than	I
	‘Eric is taller than I.’

The only feature unique to comparing quantities is that a quantifier or quantified noun is used. The compared value is usually given as mu’i ‘many’, though ha’i ‘some (plural)’ is also acceptable. The standard can be either a simple individual (64, 65, 67) or a quantified individual (66). Note that in (64) and (65) the entire complex quantifier expressions ba’ic ’i-mu’ i ‘more’ and maasma mu’ i ‘as many as’, respectively, have been floated from the restrictions. In (66) and (67), the entire construction, including the standard of comparison, may precede the auxiliary, indicating that it is one large constituent. That is, the particle mam ‘than, as’ is not a conjunction introducing a clause with ellision. If it were introducing a clause, the standard of comparison would have to be sentence final.

(64)	Homer’a-tba’ic ’i-mu’iha-huuheg pas-tiil mam
	Homeraux-pfin.frontpr-many3p-eat(pf)detpiethan
	’aapi.
	you
	‘Homer ate more pies than you.’

(65)	B-a-ñ’absha’imaasma mu’is-ha-maacheg kekel
	nr-aux-1 snrat.alllikemanystat-3p-knowdetp,man
	mam heg ‘o’oki.
	thandetp,woman
	‘I know just as many men as women.’

(66)	Ba’ic ’i-mu’i mam hetasp hemajkam’a-t’amdada.
	in.frontpr-manythanfivepersonaux-pffrcome(p)
	‘More than five people came.’

(67)	Ba’ic ’i-mu’i heg kekel mam heg ‘o’oki’at’ii
	in.frontpr-manydetp,manthandetp,womanaux-pfhere
	dadapiastawui.
	come(p)partyto
	‘More men than women came to the party.’

Maximum values can be given by using the negative to introduce the comparative construction, replacing the usual alternatives ba’ic ’ i ‘in front, more than’ or maasma ‘like, as’. This usually implies ‘less than’.

(68)	Pi	sha’i	mu’i	mam	hetasp	hemajkam	’a-t	’am	dada.
	not	at.all	many	than	five	person	aux-pf	fr	come(p)
	‘Not as many as five people came.’, ‘Less than five people came.’

13.6.3 Boolean Compounds

There are few boolean compounds of quantifiers, though they do exist. Cases generally involve the negation of the quantifier with pi ‘not’. Such boolean compounds are restricted to sentence initial position, and are among the few cases where word order and scope interact. In particular, note that negation is expressed twice in these examples, once with the quantifier and again following the auxiliary. This will be covered in depth in Section 13.7.8.

(69)	Pi	vees-ij	heg	‘a’al	’o	pi	hoohid	heg	John.
	not	all-part	det	p,child	aux	not	like	det	John
	‘Not all of the children like John.’

(70)	Pi	ha’i	kekel	’a-t	pi	’am	dada	piasta	wui.
	not	some	p,man	aux-pf	not	fr	come(p)	party	to
	‘No men came to the party.’

Even though bare nouns can be interpreted as indefinite, they cannot be negated like a quantifier can. In such cases, the negation applies to the type of individual, not the quantity.

(71)	Pi	hemajkam	’a-t	pi	’am	dada	piasta	wui.
	not	person	aux-pf	not	fr	come(p)	party	to
	‘A non-person didn’t come to the party.’ (*‘No person came to the party.’)

13.6.4 Partitives

Partitives are distinguished by a change in word order and, with some quantifiers, an additional morpheme. While typically quantifiers occur between determiner and noun, in partitives the quantifier appears outside the determiner. Vees ‘all’, mu’ i ‘many’, and ha’ i ‘some (plural)’ also take the suffix -(i)j.⁷ Recall that the default determiner is not expressed when it would otherwise be sentence initial; in partitives, since the determiner follows the quantifier, the determiner is present.

(72)	Vees	‘o’oki	’o	’e-vaila.
	all	p,woman	aux	ana-dance
	‘All women dance.’

(73)	Vees-ij	heg	‘o’oki	’o	’e-vaila.
	all-part	det	p,woman	aux	ana-dance
	‘All of the women are dancing.’

Partitives can float, just like other quantifiers (74, 75). These cases are unambiguous where the quantifier takes a suffix, but otherwise they are ambiguous. Mu’ i in (74) is unambiguously non-partitive, because the partitive form of the quantifier, mu’i j, can also float (75). A quantifier that does not have a distinct partitive form, such as gook ‘two’, does not provide any clues about partitivity when floated.

(74)	Mu’i	’a-ñ	ha-ñeid	heg	ceceoj.
	many	aux-1 s	3p-see	det	p,boy
	‘I see many boys.’

(75)	Mu’-ij	’a-ñ	ha-ñeid	heg	ceceoj.
	many-part	aux-1 s	3p-see	det	p,boy
	‘I see many of the boys.’

13.6.5 Exception Phrases

All exception phrases I have elicited have been built off of the partitive construction, though there may be other patterns available in the language. The exception phrase is introduced with shaba ‘but’, and appears at the end of the quantified expression. The entire phrase can be placed before the auxiliary, showing that the quantified expression and the exception phrase are a single constituent. Exception phrases can be added to expressions lacking a specified restriction (76) or one with an explicit restriction (77). That is, in (76) the larger group may or may not contain people that are not students, but in (77) everyone under consideration is a man.

(76)	Vees-ij shaba ga’i gook ha-mamshcamdam’a-t’amñei.
	all-partbutonlytwo3p-p,studentaux-pffrsing(pf)
	‘All of them except two students sang.’

(77)	Vees-ij heg kekel shaba pi heg John’a-t’amdada
	all-partdetp,manbutnotdetJohnaux-pffrcame(p)
	piastawui.
	partyto
	‘All of the men except John came to the party.’

13.6.6 Proportional Quantities

The only solid proportional quantity expression in the language appears to be the quantifier ’ed:a hukkam ‘half ’ (Section 13.5.1). Nevertheless, expressions referring to sub-quantities of a larger quantity can be expressed in a few different ways. Such constructions always indicate an exact number, not a mere proportion. That is, expressions like gook heg vaik ‘two of three’ refer to two objects out of three, and cannot be used for proportionally similar amounts like ‘four of six’ or ‘twenty of thirty’.

Such expressions follow a partitive pattern containing two quantifiers. The numerator is the initial, partitive quantifier; and the denominator is expressed within the restricting determiner phrase.

(78)	Gook	heg	vaik	pas-tiil	hikkmiaka	’a-n-t	ha-huu.
	two	det	three	pie	p,slice	aux-1 s-pf	3p-eat(pf)
	‘I ate two of three pie slices.’

Another strategy is to express the larger quantity via the adposition amjed: ‘from, out of’. This adpositional phrase is attached to the end of the quantified expression.

(79)	Veevkam	‘a’al	’ab	vest-maam	’amjed:	’a-t	’am	ñei.
	seven	p,child	nr	ten	from	aux-pf	fr	sing(pf)
	‘Seven children out of ten sang.’

13.7 Selected Topics

13.7.1 Type (2) Quantifiers

Type (2) quantifiers are possible in the language to the extent the vocabulary is present to create them. There is a specific word go’ol ‘different’ (80, 81), but no specific words for ‘each’ or ‘same’. ‘Same’ is expressed using the demonstratives, thus are actually deictic references to a particular individual (82). There are no special constructions or patterns for this kind of quantification.⁸

(80)	Vees	ceceoj	’o	go’ol	’uvi	s-hoohid.
	all	p,boy	aux	different	girl	stat-like
	‘All the boys like a different girl.’

(81)	Go’ol	kiik	c-’ed:	’o	kii	’am	kiihim	c-’ed:.
	different	p,house	unposs-in	aux	live	fr	town	unposs-in
	‘They live in different houses in the town.’

(82)	Vees-ij	heg	ceceoj	’o	s-hoohid	hega’i	’uvi.
	all-part	det	p,boy	aux	stat-like	that	girl
	‘All of the boys like that (the same) girl.’

13.7.2 Distributive Numerals and Binominal Each

Pima lacks an equivalent of English ‘each’, but it has a distinct morphological distributive plural pattern used with quantifiers and nouns. While collective plurality is indicated by reduplication, distributive plurality is indicated by a form of ‘double reduplication’ (with a lot of complicating phonology, see examples in Table 13.2). While some quantifiers have distinct distributive forms, the extent of this pattern is still unknown.

Table 13.2 Examples of singular, collective, and distributive forms

Distributive forms of the cardinal quantifiers get translated as groups of the base value.

(83)	Go’ogok	kekkel	’a-t	’am	dada.
	two(dist)	dist,man	aux-pf	fr	come(p)
	‘The men came in pairs/twos.’

(84)	Vavaik	kekkel	’a-n-t	’am	ha-naam.
	three(dist)	dist,man	aux-1 s-pf	fr	3p-meet
	‘I met the men three at a time.’

The same meanings can be achieved using distributive adverbs. These are constructed by suffixing -pa to one of the cardinal numbers.

(85)	Ha-mamshcamdam	’a-t	gook-pa	’e-vaav.
	3p-p,student	aux-pf	two-dist	ana-line
	‘The students lined up in pairs/twos.’

13.7.3 Mass Quantifiers and Noun Classifiers

Vees ‘all’ and ’ed:a hukka m ‘half’ can quantify either count or mass nouns (86, 87). Most if not all other quantifiers are strictly reserved for either count or mass. The three strictly mass quantifiers I am aware of are he’e s ‘how much’ and the re-purposed adjectives ge’ e ‘big, a lot, much’ and al ha’a s ‘little, a little’. All others quantify count nouns. For example, mu’i ‘many’ can modify a count noun (88), but not a mass noun (89). The correct way to express ‘much’ with a mass noun is with ge’e ‘big, a lot, much’ (90).

(86)	vees	kiiki
	all	p,house
	‘all houses’

(87)	vees	shuudagi
	all	water
	‘all the water’

(88)	mu’i	kiiki
	many	p,house
	‘many houses’
(89)	*mu’i	shuudagi
	many	water
	‘many/much water’ (intended)

(90)	ge’e	shuudagi
	big	water
	‘a lot of water’

The language does not seem to have any noun classifiers, though there are plenty of container and measure words. These expressions immediately follow the quantifier. The container/measure word can float along with the quantifier (91–94). Container words are treated like count nouns and appear in the singular or plural form as appropriate (91, 92); measure words are treated like mass nouns and are singular (93, 94).

(91)	John	’a-t	hetasp	haha’a	ha-’ii	heg	navait.
	John	aux-pf	five	p,bottle	3p-drink(pf)	det	beer
	‘John drank five bottles of beer.’

(92)	Oreos	’a-tt	gook	kokstal	ha-nolav.
	Oreos	aux-1p:pf	two	p,bag	3p-buy(pf)
	‘We bought two bags of Oreos.’

(93)	Hetasp	pisal	’a-n-t	ha-nolav	heg	pilkañ.
	five	pound	aux-1 s-pf	3p-buy(pf)	det	wheat
	‘I bought five pounds of wheat.’

(94)	Gi’ig	novi	cev	’a-t	tatcua	heg	vijna.
	four	arm	long	aux-pf	want	det	rope
	‘I need four arm-lengths of rope.’

13.7.4 Existential Constructions

There are two common existential constructions. One is a typical intransitive sentence with one of a small number of existential verbs. The most semantically vague of these is ha’icu g ‘exist’, but others include shuudagi ‘exist (liquid)’, kuubs ‘exist (smoke)’, and kaac ‘exist (lots of small particles)’. If the subject is inanimate, it is common to leave the verb implied (95), but a verb is almost always present with an animate subject (97). Often, the subject will take the determiner ge ‘a certain’, but this is not always the case.

(95)	Ge	hahag	’o	’am	miish	veeco	(ha’icug).
	certain	leaf	aux	fr	table	under	exist
	‘There is a leaf under the table.’

(96)	Ge	‘o’od	’o	’am	kooba	c-’ed:	kaac.
	certain	sand	aux	fr	cup	unposs-in	exist
	‘There is sand in the cup.’

(97)	Gook	kekel	’o	kii	c-’ed:	*(ha’icug).
	two	p,man	aux	house	unposs-in	exist
	‘There are two men in the house.’

The other pattern is to convert the noun denoting the individual into a predicate adjective (98). The noun is almost always in the plural form, though in some elicited examples the singular has been found. This construction always has a locative phrase in it, and there does not appear to be any subject.⁹ That the denominal adjective is the predicate can be shown by affixing tense/aspect morphology to it (99).

(98)	Kui	veeco	’o	s-totobi-g.
	tree	under	aux	stat-p,rabbit-adj
	‘There are rabbits under the tree.’

(99)	Kui	veeco	’o	s-totobi-g-kahim.
	tree	under		aux	stat-p,rabbit-adj-pst:cont
	‘There used to be rabbits under the tree.’

The same derivational suffix is used to derive characteristic adjectives from nouns.

(100)	S-kui-g	’o	heg	’oid:ag.
	stat-tree-adj	aux	det	field
	‘The field is tree-y [full of trees].’

(101)	S-jevd:a-g	’o	heg	cevho	kii.
	stat-dirt-adj	aux	det	gopher	house
	‘Gopher’s house is dirt-y [made of dirt].’

Despite the adjectival morphology, the underlying nouns still behave as nouns in some respects: they can be quantified and they can antecede a pronoun (102, 103). In these constructions, the quantifier is in the typical quantifier float position before the predicate adjective.

(102)	M-o	mu’i	s-totobi-g	kui	veeco.	N-a-p	ha-ñeid?
	fr-aux	many	stat-p,rabbit-adj	tree	under	q-aux-2 s	3pl-see
	‘There are lots of rabbits under the tree. Do you see them?’

(103)	M-o	vees	s-totobi-g	kui	veeco.	N-a-p	ha-ñeid?
	fr-aux	all	stat-p,rabbit-adj	tree	under	q-aux-2 s	3pl-see
	‘All of the rabbits are under the tree. Do you see them?’

Both types of existential constructions behave like typical intransitive predicates. There are no significant differences regarding negation or question formation.

13.7.5 Floating Quantifiers

All d-quantifiers can be floated to pre-verbal position, forming a loose constituent with the verb, as discussed in depth by Munro (1984). The rules for resolving which noun the quantifier was floated from get complex, and the patterns exhibited by my consultant differ from those of Munro’s consultant. It should be noted that Munro’s consultant and mine were from different generations and different communities, so this is likely a dialectal difference. Indeed, it seems likely that my consultant’s dialect simply has stricter resolution rules, not a completely different set of them.

In the case of a simple transitive sentence, the floated quantifier is interpreted with the object.

(104)	‘O’oki	’a-t	vees	’i	ha-daad:sh	heg	‘e-’a’al.
	p,woman	aux-pf	all	inc	3p-p,make.sit	det	ana-p,child
	‘The women sat all their children down.’
	*‘All the women sat their children down.’

Munro (1984) reports that her consultant would allow the quantifier to associate with the subject, if the object was semantically incompatible. For example, in a case where the quantifier requires a plural noun, but the object is singular, the quantifier can associate with a plural subject. My consultant rejects such sentences.

(105)	%Hegam	ceceoj	’o	vees	ñeid	heg	Alice.
	those	p,boy	aux	all	see	det	Alice
	‘Those boys all saw Alice.’ (intended)

Either object of a ditransitive verb can float a quantifier. There are differences between Munro’s consultant and mine here as well, though there are also some telling similarities. The main difference is that Munro’s consultant allowed both objects to float quantifiers at the same time. Both move to the standard pre-verbal position, and which object each quantifier is associated with is determined by linear precedence: the first quantifier quantifies over the first object, the second quantifier the second object. This is irrespective of which is the direct object or indirect object. Word order is free amongst arguments, so this is a pure linear order issue.

(106)	Rina	’a-t	gook	ha’i	ha-maa	heg	‘e-’o’’ohan	hegam
	Rina	aux-pf	two	some	3p-give(pf)	det	ana-p,book	those
	mamakai.
	p,doctor
	‘Rina gave two of her books to some of the doctors.’

Table 108

My consultant only permits one quantifier to be floated, but it can float from either object. There is a preference for the quantifier to modify the linearly closest object, but this is only a preference, not a requirement. Thus, though there are different syntactic patterns between these two varieties of Pima, there appears to be a core similarity at work.

(108)	Heñ-nawpuj	’a-t	ha’i	ha-maa	hegam	ceceoj	heg	‘o’’ohan.
	1 s-p,friend	aux-pf	some	3p-give(pf)	those	p,boy	det	p,book
	‘My friends gave the books to some of the boys.’ (preferred) or
	‘My friends gave some of the books to the boys.’

Munro reports that intransitive verbs allow quantifier float from their subject. She makes no reference to distinctions between different types of intransitive verb. With my consultant, whether or not floating is possible depends on the lexical class of the verb. There is a three way distinction: Unaccusative verbs allow float from the subject (109, 110). Verbs with incorporated objects allow float from that underlying object (111, 112).¹⁰ Unergative verbs generally do not permit floating at all, though the data here are noisy: occasionally my consultant judged (113) and (114) as acceptable.

(109)	Kekel	’a-t	gook	’ii	dada.
	p,man	aux-pf	two	here	come(p)
	‘Two men arrived.’

(110)	Gogogs	’a-t	gook	ko’ok.
	p,dog	aux-pf	two	die(p,pf)
	‘Two dogs died.’

(111)	Kekel	’o	gook	kii-t.
	p,man	aux	two	house-make
	‘The men are building two houses.’

(112)	‘O’oki	’o	gook	paan-t.
	p,woman		aux	two bread-make
	‘The women are making two loaves of bread.’

(113)	*Gogogs	’o	gook	tototk.
	p,dog	aux	two	p,bark
	‘Two dogs are barking.’ (intended)

(114)	*‘A’al	’o	gook	shoañ.
	p,child	aux	two	cry
	‘Two children are crying.’ (intended)

There is one further difference in floated quantifiers between the two consultants: Munro reports that quantifiers can be floated from possessors of objects. My consultant can only interpret such sentences with the quantifier modifying the object, not the object’s possessor. In the following example, the object is singular and therefore not an acceptable restriction for vees ‘all’. Munro’s consultant accepted it, but mine does not.

(115)	%Vees	ñei	’a-n-t	heg	heñ-nawpuj	ha-maakai-ga.
	all	see(pf)	aux-1 s-pf	det	1 s-p,friend	3p-doctor-poss
	‘I saw the doctor of all my friends.’ (intended)

The patterns observed from Munro’s consultant and my own differ in many crucial respects, but not in random ways. It appears that my consultant’s variety has a reduced syntactic distribution of floated quantifiers by disallowing them in all contexts where the restriction is not a syntactic object in some sense. This extends into the intransitive domain, where arguments that are standardly accepted to be object-like (the subject of unaccusatives and incorporated nouns) permit floating, but arguments that are more subject-like (the subject of unergatives) do not.

13.7.6 Bare Quantifiers as Arguments

It is possible that all quantifiers can be used as bare arguments, though the data are not convincing. The issue is that virtually any argument can be zero pronominalized; thus it could be difficult if not impossible to determine if the quantifier is the sole element of the argument, or if it is modifying a silent pronoun. It is worth pointing out, though, that adjectives and postpositional phrases cannot be stranded by zero pronominalization. Thus, if quantifiers can be, the rules for zero pronominalization apply different to them than any other adnominal modification. The two examples below are about as clear evidence as you can find. Still, (116) could alternatively be analyzed as ‘They are two women’ (with quantifier float), and (117) could be ‘She kissed them all’ (with pro-drop):

(116)	Gook	’o-d:	‘o’oki.
	two	aux-cop	p,woman
	‘Two are women.’

(117)	M-a-t	vees	ha-cindat.
	fr-aux-pf	all	3p-kiss
	‘She kissed everyone.’

13.7.7 Bare Quantifiers as Predicates

In certain cases, a quantifier can serve as the main predicate. Cardinal numbers cannot serve as a predicate on their own. However, they are acceptable with the copula. This suggests they are functioning like a predicate nominal in these cases. Verbal, adjectival, and adpositional predicates do not co-occur with the copula.

(118)	*’Iidam	’o	gook.
	these	aux	two
	‘They are two (in number).’ (intended)

(119)	‘Iidam	’o-d:	gook. (Mark 16:12; Papago and Pima Translators, et al. 1975)
	these	aux-cop	two
	‘They are two (in number).’

The partitive forms veesij ‘all of ’, ha’i j ‘some of ’, and mu’i j ‘many of ’ can be predicates, but the non-partitive forms cannot. Evidence that they are predicates comes from the presence of tense/aspect morphology. (It is worth pointing out that quantifiers as main predicates have only been observed under direct elicitation. They have not been spontaneously produced.)

(120)	Kui	veeco	’o	mu’i-j-kahim	heg	totobi.
	tree	under	aux	many-part-pst:cont	det	p,rabbit
	‘There were many rabbits under the tree.’

(121)	Totobi	’o	’am	vees-ij-kahim	kui	veeco.
	p,rabbit	aux	fr	all-part-pst:cont	tree	under
	‘All the rabbits were under the tree.’

Mu’i j ‘many of’ can serve as a predicate by itself (122). Veesij ‘all of’ is ungrammatical without a locative phrase (123 vs. 121). The acceptability of ha’ij ‘some of’ in such a context is unknown.

(122)	Totobi	’o	mu’-ij.
	p,rabbit	aux	many-part
	‘The rabbits are many.’

(123)	*Kekel	’o	vees-ij.
	p,man	aux	all-part
	‘The men are all.’ (intended)

13.7.8 Scope Ambiguities

When more than one quantificational expression appears in a single clause, they can usually scope in either order. There is no correlation between scope and word order, part of speech, or floating status. There may be some preferences for one reading over another in certain environments, but these have not been successfully teased apart at this point.

(124)	Hema	’ali	’a-t	vees	ha-ñeid	heg	‘o’’ohan.
	a	child	aux-pf	all	3p-see	det	p,book
	‘A child read all the books.’ (some > all, all > some)

(125)	Hema	‘o’idam	’a-t	gook	ha-kokkeda	heg	huahi.
	a	hunter	aux-pf	two	3p-kill(p)	det	p,deer
	‘A hunter killed two deer.’ (some > two, two > some)

While this is the general case, there are times when scope is judged to be unambiguous. It is not clear when or why this is true. In the following case (126), each hunter killed a separate deer. The reading where a single deer is killed by all the hunters collectively is reported to be impossible.

(126)	Vees-ij	heg	‘o’’idam	’a-t	hema	mua	heg	huai.
	all-part	det	p,hunter	aux-pf	a	kill	det	deer
	‘All of the hunters killed a deer.’

Quantified subjects scope over or below sentential negation based on relative word order. When the quantifier is to the left of negation, it scopes over the negation (127). When the quantifier is to the right, negation scopes over it. This is regardless of whether the quantifier is floated (129) or not (128).

(127)	Hema	keli	’a-t	pi	’am	jivia	piasta	wui.
	a	man	aux-pf	not	fr	arrive	party	to
	‘A man didn’t come to the party.’ (some > not)

(128)	Pi	’a-t	’am	jivia	heg	hema	keli	piasta	wui.
	not	aux-pf	fr	arrive	det	a	man	party	to
	‘No man came to the party.’ (not > some)

(129)	Pi	’a-t	’am	hema	jivia	heg	keli	piasta	wui.
	not	aux-pf	fr	a	arrive	det	man	party	to
	‘No man came to the party.’ (not > some)

If the quantified noun phrase needs to scope under negation, but needs to be to the left of it for discourse reasons, both the quantifier and the sentence are negated (130). The quantifier cannot be modified by negation if it is already under the scope of sentential negation (131).

(130)	Pi	hema	keli	’a-t	pi	’am	jivia	piasta	wui.
	not	a	man	aux-pf	not	fr	arrive	party	to
	‘No man came to the party.’

(131)	*Pi	’at	’am	jivia	heg	pi	hema	keli	piasta	wui.
	not	aux-pf	fr	arrive	det	not	a	man	party	to
	‘No man came to the party.’ (intended)

Quantified objects are reported to be scope ambiguous when they occur before negation.

(132)	Hema	keli	’a-ñ	pi	ñeid.
	a	man	aux-1 s	not	see
	‘I didn’t see a man.’ (some > not, not > some)

There is still much work to be done before scope in Pima is understood, but the above gives some idea of the patterns observed thus far.

13.7.9 Only

The syntax and semantics of ‘only’ are not well understood at this point. There are two apparent morphemes that translate as ‘only’. ’Absh ‘just, only’ distributes as an adverb and restricts predicates. This includes verbal and nominal predication.

(133)	’Iiya	’a-c	’absh	dad:he.
	here	aux-1p	just	p,sit
	‘We just sat here. (We did not do anything else.)’

(134)	D-a-ñ	’absh	‘o’odham.
	cop-aux-1 s	just	human
	‘I’m only human.’

Individuals are restricted via a series of words that all seem to include a suffix -a’ i: va’ i, ma’ i, and ga’ i. There may be others. These words are typically found very close in front the verb. They can modify either the subject (135) or object (136, 137); and the noun modified can precede (137) or follow (136) the particle.

(135)	Hega’i	’uvi	’o	va’i	cikpan.
	that	young.woman	aux	only	work
	‘Only that young woman is working (nobody else).’

(136)	M-a-ñ	ga’i	ñeid	heg	John.
	fr-aux-1 s	only	see	det	John
	‘I saw only John.’

(137)	Pam	’a-ñ	ma’i	s-maac.
	Pam	aux-1 s	only	stat-know
	‘I only know Pam.’

Ga’ i ‘only’ is occasionally found at the beginning of a determiner phrase, before demonstratives (138) or quantifiers (139). The other two forms do not appear inside determiner phrases.

(138)	ga’i	hegam	ceceoj
	only	those	p,boy
	‘only those boys’

(139)	ga’i	hemako	ceoj
	only	one	boy
	‘only one boy’

The most obvious analysis of these forms would be so segment off the -a’i as meaning ‘only’, attaching to common adverbial particles: va a certain future, ’am the ‘behind’ deictic, and ge the specific indefinite, respectively. This may be true; however, the words meaning ‘only’ appear at times in sentences were the particles would not normally be found, suggesting they may have an independent existence. For example, ge does not appear before demonstratives, but ga’i appears before one in (138); and the aspectual particle va only appears in futures and in certain modal contexts, but appears in an imperfective sentence in (135). The details of distribution have not been worked out yet.

There is no clear difference in meaning between any of these three forms. All three can modify a subject or an object (135 vs. 136). However, they are not interchangeable in all contexts. For example, ga’ i and ma’i are judged interchangeable in (140) and (141), but va’i in the same context is unacceptable (142). The conditions governing the distribution are as yet unknown.

(140)	Pam	’a-ñ	ga’i	s-maac.
	Pam	aux-1 s	only	stat-know
	‘I only know Pam.’

(141)	Pam	’a-ñ	ma’i	s-maac.
	Pam	aux-1 s	only	stat-know
	‘I only know Pam.’

(142)	*Pam	’a-ñ	va’i	s-maac.
	Pam	aux-1 s	only	stat-know
	‘I only know Pam.’ (intended)

13.8 Conclusion

The description of quantifiers in Pima presented in this paper is far from a complete accounting of the patterns in the language, but it shows that the topic is a rich one, with much still to explore. Nevertheless, it seems clear that quantification is complex and productive.