amerindia n°21, 1996

Grammatical Hispanisms in Amerindian and Austronesian Languages

The Other Kind of Transpacific Isoglosses

Thomas Stolz

Ruhr-University/Bochum, Germany

In the present paper^[1], I seek to demonstrate that what up until recently has appeared to be a phenomenon restricted to Middle America rather turns out to be a common trait of quite a number of languages on both shores of the Pacific Ocean. Irrespective of the differences of their individual genetic, areal, and typological affiliation, many languages spoken in Middle America, South America, in the Philippines and Marianas share a set of lexemes and grammemes borrowed from Spanish either during the age-long colonial rule of Spain or, at least for what concerns the Americas, also in post-colonial times.

Discounting the many correspondences in the selection of borrowed lexical items, what strikes us most is the fact that Austronesian and Amerindian languages display similar preferences when it comes to borrowing function words from Spanish, the more so since there still prevails a prejudice claiming that function words are less likely to be borrowed at all. However, the evidence I have been able to gather so far does not support this hypothesis. Rather, it seems that function words count among the most frequent, wide-spread, and consistent borrowings in the 41 languages of my present sample^[2] (for a detailed list of the languages and a spell-out of the abbreviations used, cf. appendices 1-2).

As (1) is meant to suggest in a primitive version of set-theoretical reasoning, the inventory of function words (= F) of languages from Middle America (= L_(M)), South America (= L_(S)), the Philippines and Oceania (= L_(P)) does at least partially overlap with that of Spanish (= L_(E)). If a given function word shows up in the inventories of minimally one language of each area, then this is an instance of a circumpacific isogloss (= C). There are also other, less extended, types of isoglosses to be observed in my present sample, cf. (2) (with x = function word). However, in the remainder of my presentation, I will exclusively deal with the circumpacific type because, as far as theory is concerned, it is the most interesting of all isoglosses.

(1) Overlapping Sets

(2) types of Isoglosses

(a) intraareal

x ' F(L_(Mi)) + F(L_(Mj)) [+F(L_(E))] Middle American

x ' F(L_(Si)) + F(L_(Sj)) [+F(L_(E))] South American

x ' F(L_(Pi)) + F(L_(Pj)) [+F(L_(E))] Pacific

(b) interareal (binary)

x ' F(L_(M)) + F(L_(S)) [+F(L_(E))] Pan-/Transamerican

x ' F(L_(M)) + F(L_(P)) [+F(L_(E))] Northern Transpacific

x ' F(L_(S)) + F(L_(P)) [+F(L_(E))] Southern Transpacific

x ' F(L_(M)) + F(L_(S)) + F(L_(P)) [+F(L_(E))] Circumpacific

For a start, consider sentence (3) which is an example of the every-day use of the Spa-nish adversative conjunction pero ‘but’ to be encountered in all European and non-European varieties of that language. Now, pero is not confined to Spanish and its regional varieties, rather it is also an integral part of the function-word inventories of languages like e.g. Nahuatl in Middle America, cf. (4), Guaraní in South America, cf. (5), and Tagalog in the Philippines, cf. (6). As one can easily see, pero is used here with much the same adversative function as in the donor-language Spanish i.e. as a translational equivalent of English but. Therefore, according to the definition given in (1), the borrowing of pero constitutes a circumpacific isogloss.

(3) Spanish

sí sabe español pero no quiere hablar contigo ahora

sure know:3Sg Spanish but Neg want:3Sg speak:Inf with:2Sg now

‘Of course he knows Spanish, but he doesn’t want to chat with you right now.’

(4) Nahuatl (Preuss & Ziehm 1971, 68)

pero yéual uksé in tamasol

but 3Sg.Emp one.more Def toad

‘But that was another toad.’

(5) Guaraní (Gregores & Suárez 1967, 186)

pero ko'ero a- -hà yeí ta la mediko rená pe

but tomorrow 1Sg.Act -go again Fut Def doctor place Loc

‘But tomorrow I’ll go again to see the doctor.’

(6) Tagalog (Schachter & Otanes 1972, 544)

hindi namin ma -ga -gawa ngayon pero ga -gaw -in namin bukas

Neg 1Pl Apt -Red -do today but Red -do -OF 1Pl tomorrow

‘We can’t do it today, but we’ll do it tomorrow.’

As a matter of fact, pero is far from being the only case of a circumpacific isogloss, not to mention function-word borrowing in general^[3]. For the latter, consider (7) which is a list of Spanish-derived function words in a dialect of Yucatec Maya as spoken in some villages of Quintana Roo, Mexico. Even with its 29 borrowed function words, however, Yucatec is only 4th best in Middle America. Table (8) covers 19 Middle American languages and gives the total number of borrowed Spanish function words for each of them^[4]. The Middle American average is 15 borrowed function words per language. In addition, table (8) gives the sub-total of items which belong to the regional core C_(M) of borrowed function words with an average of slightly more than 6 per language.

(7) Yucatec

antes ‘before’; bueno ‘well’, cada ‘each’; casi ‘almost’; como ‘since; as’; de Yucatec attribute marker/depredicativizer < Spanish vestido de rojo ‘dressed in red’; desde Yucatec ‘from, till’ < Spanish ‘from’; después ‘then, after that’; este hesitation; hasta ‘till; even’; más ‘more’; mientras ‘while’; mismo ‘self; even’; o ‘or’; pero ‘but’; por ‘by; because of’; porque ‘because’; pues ‘well then’; puro ‘only’; que ‘that’; si ‘if’; sin ‘without’; sino ‘but rather’; sí ‘of course’; siempre ‘always’; tal vez ‘perhaps’; todavía ‘still’; y ‘and’; ya ‘already’

(8) Size of Inventories of Borrowed Function Words in Middle America

Total members of the regional core C_(M)

1. Popoloca 37 9

2. Zapotec 36 11

3. Otomí 34 10

4. Pipil 29 10

Yucatec 29 9

5. Nahuatl 24 11

6. Totonac 21 9

7. Mam 19 5

8. Popoluca 14 7

9. Tzutujil 11 6

Cora 11 6

10. Yaqui 8 4

11. Chinantec 7 5

Ch'ol 7 3

12. Mixe 3 3

Mixtec 3 3

13. Garífuna 2 2

Jacaltec 2 1

14. Mazatec 1 1

19 287 115

The regional core is made up of the sub-set of borrowed function words with the highest frequency and widest distribution within a given area. As for Middle America, any item occurring in more than 3 languages of the sub-sample has been included in C_(M). Accordingly, table (9) lists the 12 members of C_(M) and their distribution over the languages of my Middle American sub-sample with decreasing frequency from top to bottom for the number of languages in which the borrowed item has been attested and from left to right for the number of borro-wings attested in individual languages. Note that among the 6 most frequent borrowings, i.e. those with a frequency of 10 and more, there is not a single preposition. The majority of cases consists of conjunctions and discourse particles with pero being the most prominent among them. On top of that, there is, in (9), an implication requiring that if a language has borrowed a preposition from Spanish, then it always also has borrowed conjunctions and discourse par-ticles from Spanish, whereas the reverse does not hold good.

(9) Members of C_(M) and their Distribution over the Middle American

Sub-Sample

pero

porque

bueno

más

sino

como

para

antes

con

CH'

ch'

115

For brevity’s sake, consider South America now. A language that has been described as almost free of grammatical Hispanisms is Aymara, for instance. Nevertheless, one may easi-ly come across the 7 borrowed function words under (10) again including pero. All the other languages of my South American sub-sample have borrowed more function words from Spanish as results from table (11). This time, the average of borrowed function words per language is about 19 items. The regional core C_(S) for South America is given in table (12). There can be no doubt that C_(M) and C_(S) contain a number of common elements which, according to (2), may be labeled constituents of transamerican or panamerican isoglosses. As is also the case with C_(M), pero is again the most prominent item. Moreover, the 5 most frequent items are identical in both core-sets with only a change in the ranking of o ‘or’ and porque ‘because’. In addition, there is no preposition among the members of C_(S). In this way, the implication deduced from the Middle American sub-sample is corroborated by the South American data.

(10) Aymara

antes ‘before’; a veces ‘at times’; bueno ‘well’; entonces ‘then’; pero ‘but’; sino ‘but rather’; y ‘and’

(11) Size of Inventories of Borrowed Function Words in South America

Total members of the regional core C_(S)

1. Cochabamba Quechua 39 7

2. Guaraní 34 7

3. Ayacucho Quechua 18 6

4. Inga 16 6

5. Mapuche 14 2

6. Huánuco Quechua 8 2

7. Aymara 7 4

7 136 34

(12) Members of C_(S) and their Distribution over the South American Sub-Sample

		pero	o	porque	ni	bueno	sino	más que	antes
7	GU	gu	gu	gu	gu	gu	gu	gu
7	CQ	cq	cq	cq	cq	cq	cq		cq
6	AQ	aq	aq	aq	aq			aq	aq
6	IN	in	in	in	in		in	in
4	AY	ay				ay	ay		ay
2	HQ	hq	hq
2	MP	mp				mp
34	7	7	5	4	4	4	4	3	3

Grossly speaking, one encounters much the same situation in Oceania. Take e.g. the Philippinian language Hiligaynon. In (13), I enumerate the 14 borrowed Spanish function words attested for Hiligaynon. As one may already have expected, pero is once again a member of the inventory. Table (14) clearly shows that there are several other languages in the same area which score much higher than Hiligaynon with regard to function-word borrowing. The Austronesian average of 10 borrowings per language is comparatively low - and so is the average of just 3 items per language belonging to the regional core C_(P). In table (15), the 8 members of C_(P) are presented.

(13) Hiligaynon

antes ‘before’; demasiado ‘too (much)’; entonces ‘then’; fuera ‘outside; except’; hasta ‘till’; más ‘more’; más que Hiligaynon ‘even if’ < Spanish ‘although’; medio ‘half-’; mismo ‘self’; o ‘or’; para ‘for’; pero ‘but’; seguro Hiligaynon ‘perhaps’ < Spanish ‘surely’; siempre Hiligaynon ‘surely’ < Spanish ‘always’

(14) Size of Inventories of Borrowed Function Words in Austronesia

Total members of the regional core C_(P)

1. Chamorro 32 4

2. Bikol 26 8

3. Cebuano 2 5

4. Tagalog 19 6

5. Hiligaynon 14 7

6. Tagbanwa 10 6

Kapampangan 10 5

7. Manobo 6 3

8. Batak 5 2

9. Mansaka 4 2

10. Sambal 3 3

Pangasinan 3 2

11. Agta 2 2

Negrito 2 2

Rapanui 2 2

15 160 59

(15) Members of C_(P) and their Distribution over the Austronesian Sub-Sample

		o	más que	pero	porque	para	antes	más	fuera
8	BI	bi	bi	bi	bi	bi	bi	bi	bi
7	HI	hi	hi	hi		hi	hi	hi	hi
6	TA	ta	ta	ta	ta	ta		ta
6	TG	tg	tg	tg	tg		tg		tg
5	CE	ce		ce	ce	ce		ce
5	KA	ka		ka	ka		ka		ka
4	CM		cm			cm	cm		cm
3	MB			mb	mb	mb
3	SB	sb	sb	sb
2	BT		bt			bt
2	MK		mk		mk
2	PG	pg					pg
2	AG	ag	ag
2	NE	ne	ne
2	RP	rp		rp
59	15	11	10	9	7	7	6	5	4

With the exception of the least frequent item, i.e. fuera ‘except’, and the highly disputable case of para ‘for’, which according to some authors does not seem to be a proper case of borrowing after all, there again is a scarcity of prepositions among the borrowed function words. Anyway, the implication according to which every borrowed preposition implies the borrowing of a conjunction or discourse particle once more is fulfilled. The bulk of the borrowed function words comprises conjunctions and discourse particles many of which do also form part of the cores C_(M) and C_(S), respectively. What distinguishes C_(P) from the two American core sub-sets is, on the one hand, the unusually high frequency of the concessive conjunction más que ‘although’, and, on the other hand, the change in ranking position of o and pero. In spite of these minor changes, the general picture remains unaltered i.e. the inventories of C_(M), C_(S), and C_(P) all resemble one another.

Before proceeding with the discussion of the isoglosses proper, a short excursus seems to be in order. The excursus takes up another issue connected with the homologies in the borrowing behaviour of Austronesian and Amerindian languages. One tentative explanation for the borrowing of function words in the first place may be labeled the gap hypothesis. Whoever may favour this hypothesis assumes that the borrowing of Spanish function words would have been necessary because the autochthonous languages lacked functional equivalents of the Spanish particles, conjunctions, and sundry grammatical elements. Following this line of reasoning, the borrowed items would have filled in structural gaps in the systems of the recipient languages.

However, I cannot subscribe to this point of view especially because there seems to be no reason for assuming pre-colonial structural gaps if one takes into account the fact that, synchronically, there is ample evidence for the co-existence of partially or even fully synonymous autochthonous and borrowed items in one and the same grammeme inventory. One such case in point is Yucatec. As (16) shows, many of the Spanish borrowings have corresponding Yucatec equivalents (in italics); in this case, there are synonyms for 13 out of a total of 29 borrowings. Furthermore, besides the Spanish-derived pero, there is also a synonymous indi-genous adversative conjunction (chem) ba’xe’/ba’n ‘but’, as sentences (17)-(18) illustrate.

(16) Yucatec

bueno = ma'lo'b ‘well’; cada = reduplication ‘each’; como = bix túun/he'x ‘as; since’; entonces = halibe' ‘then’; hasta = tak ‘till’; o = wáah ‘or’; pero = (chem) ba'xe'/ba'n ‘but’; por = men ‘by; because of’; porque = tumen ‘because’; pues = túun ‘well then’; sin = xma' ‘without’; todavía = layli'e' ‘still’; y = ká(ah), yéetel ‘and’

(17) Yucatec (Smailus 1975: 14)

tih máan -o'b -i' pero ma' uts t ich -o'b -i'

there pass -Pl -D4 but Neg good Loc eye -Pl -Neg

‘They passed by the place, but they didn’t like it there.’

(18) Yucatec (Smailus 1975: 214)

ba'n bey -o' buka'h óotsil -o'b t -u tus -ah?

but so -D2 how=many poor -Pl Prt -Erg.3 lie -Trs.Com

‘But how many poor souls have they deceived?’

Neither is the co-existence of borrowed and inherited items restricted to Yucatec, as the examples from Aymara, (19)-(20), and Hiligaynon, (21)-(22), show, nor does it only affect the borrowing of pero, as the alternation of o and ukon ‘or’ in the Hiligaynon sentences (23)-(24) make clear. Indeed, it is a commonality of Austronesian and Amerindian languages to have such mixed pairs of synonyms. As for Middle America, the use of both members of such a pair of synonyms side by side in one and the same construction is well attested. It remains to be investigated to what extent such function-word doublets consisting of Spanish and indigenous items do also occur outside of Middle America. To keep things short, the co-existence of borrowed and inherited items in a synonym relation, in my opinion, disproves the gap hypo-thesis since instead of gaps there actually are indigenous functional equivalents. Thus, one has to look for a different explanation of the facts.

(19) Aymara (Porterie-Gutierrez 1988: 362)

pero sorro -sti wali astuturi -tajna...

but fox -Coo very keen -3Sg.Prt.Evi

‘But the fox was very keen.’

(20) Aymara (Porterie-Gutierrez 1988: 355)

sorro -sti wali astuturi -tajna -wa

fox -Coo very keen -3Sg.Prt.Evi -CFP

‘But the fox was very keen.’

(21) Hiligaynon (Wolfenden 1971: 73)

gusto ko tani mag -lakat pero naga -ulan

like 1Sg wish AF.Fin -walk but AF.Dur -rain

‘I would like to go for a walk, but it is raining.’

(22) Hiligaynon (Wolfenden 1971: 73)

mainit ang adlaw apang ang mga tawo naga -linakat sa baybay

hot Def sun but Def Pl.Hum person AF.Dur -go.Rea Loc beach

‘The sun is hot, but the people are going to the beach anyway.’

(23) Hiligaynon (Wolfenden 1971: 77)

pa -lakt -on mo si Jose o si Juan

Cau -go -OF 2Sg Def.Hum José or Def.Hum Juan

‘Will you have José go or Juan?’

(24) Hiligaynon (Wolfenden 1971: 77)

pa -lakt -on mo si Jose ukon si Juan

Cau -go -OF 2Sg Def.Hum José or Def.Hum Juan

‘Will you have José go or Juan?’

This brings us back to the main topic of my paper. Compare (4)-(6) with three more examples of pero borrowed into Middle American Totonac, cf. (25), South American Quechua, cf. (26), and Philippinian Kapampangan, cf. (27). Such an accumulation of evidence lends strong support to the assumption that there must be something special about pero.

(25) Totonac (Reid, Bishop, Button & Longacre 1968: 141)

pero quit i'ca'mâ' â'lacatin

but 1Sg go:Fut elsewhere

‘But I am going to another place.’

(26) Spoken Bolivian Quechua (Bills, Vallejo & Troike 1969: 100)

pero aycha apamu -q bende -war -qa waka aycha -ta

but meat bring -NA sell -1Sg.Ben -3Sg.Prt cow meat -Acc

‘But the meat vender sold me some beef.’

(27) Kapampangan (Mirikitani 1971: 102)

pero reng kapatad na -ng babai mi -bait la king Amerika

but Top sibling Psr.3Sg -Lig woman Ina -give=birth 3Pl Loc America

‘But his sisters were born in America.’

At this point, we are in a position to identify those isoglosses which qualify for the status of circumpacific isogloss by way of mapping the three core-sets C_(M), C_(S), and C_(P) onto each other. Once again, this is done by a layman’s application of set-theoretical principles in (28). Skipping for once those overlapping sections which involve only 2 areas, it is the core of cores - framed in bold face - which is in the centre of interest here. This core of cores consists of 4 items, viz., in alphabetical order, antes ‘before’, o ‘or’, pero ‘but’, and porque ‘because’. Put differently, exactly these 4 items count always among the most frequent cases of borrowing no matter which regional sub-sample is concerned. In fact, these 4 items are the paradigm cases of circumpacific isoglosses.

(28) Identifying the Core of Cores

To complicate things a little bit, antes, o, pero, and porque, however, are not evenly distributed among the languages of my present sample. Consider table (29): it actually is a list combinedly displaying, (a), the distribution of core-of-core members over the 41 languages put under scrutiny, and, (b), the participation of the individual languages in circumpacific isoglosses. Only 2 out of 41 languages do not participate in any of the circumpacific isoglosses; on the other hand, there are 11 languages that have their share in all 4 circumpacific isoglos-ses. It is important to note that among these 11 languages, there are representatives of all 3 sub-areas. Pero forms the most extended isogloss to which 80% of the languages belong. 68% is the score for o, 58% for porque, and 36% for antes.

(29) Synopsis of the Patterns of Distribution of C_(MSP) over individual languages

	pero	o	porque	antes	*others only*	area
4	ZA	ZA	ZA	ZA		M
4	NA	NA	NA	NA		M
4	OT	OT	OT	OT		M
4	PI	PI	PI	PI		M
4	PO	PO	PO	PO		M
4	YU	YU	YU	YU		M
4	CQ	CQ	CQ	CQ		S
4	AQ	AQ	AQ	AQ		S
4	BI	BI	BI	BI		P
4	TG	TG	TG	TG		P
4	KA	KA	KA	KA		P
3	TO	TO	TO			M
3	TZ	TZ	TZ			M
3	CO	CO	CO			M
3	CH	CH	CH			M
3	GU	GU	GU			S
3	IN	IN	IN			S
3	TA	TA	TA			P
3	CE	CE	CE			P
3	HI	HI		HI		P
2	MX	MX				M
2	HQ	HQ				S
2	SB	SB				P
2	RP	RP				P
2	PP		PP			M
2	MM		MM			M
2	MB		MB			P
2	AY			AY		S
2		YA	YA			M
2		PG		PG		P
1	CH'					M
1	MI					M
1	GA					M
1	MZ					M
1	MP					S
1		AG				P
1		NE				P
1			MK			P
1				CM		P
0					BT	P
0					JA	M
41	33	28	24	15	2

Most of what (29) depicts in an enumerative fashion is also represented on the four maps, cf. appendix 3, each of which provides for a geographical view of the four circumpacific isoglosses. The maps are somehow suggestive of a geographical correlate that accompanies the decrease in extension of the isoglosses: those languages which are spoken in the southernmost parts of South America and Oceania are the first to drop from the isoglosses as soon as these become less extended. This is the normal picture of wave-like diffusion from an innovatory center in language-change processes. However, there is, in fact, no diffusional center for the spread of borrowings in South America other than the regional varieties of Spanish themselves. For the time being, we may content ourselves with speculating that the chronology and history of language contacts, which, for instance, started comparatively late in the case of Rapanui, could be responsible for the above mentioned phenomenon.

Turning back to (29), it is possible to deduce from the list at least two implications. The first, cf. (30), states that the combined borrowing of porque and antes into a given language requires that both pero and o have also been borrowed into the same language. The second implication, cf. (31), states that if there are more than two members of the core of cores borrowed by a given language then pero is always one of them.

(30) Implication A

If a language L_(MSP) has borrowed porque ‘because’ and antes ‘before’, then it always has borrowed pero ‘but’ and o ‘or’ as well.

(31) Implication B

If a language L_(MSP) has borrowed more than 2 function words that belong to C_(MSP), then pero ‘but’ is always among the borrowed items.

To sum up, the prominent role that pero plays in function-word borrowing is under-scored by its high frequency as a member of every regional core-set, by its wide distribution in circumpacific perspective, and, last but not least, also by the nature of the above implications. In a way, pero may thus be claimed to represent the unmarked case of function-word borrowing. Since many of the other highly frequent items are also conjunctions or discourse particles, it is legitimate to render the markedness assumption more general by saying that elements which help organizing discourse form the unmarked cases of grammeme borrowing. Even for what concerns apparent instances of preposition borrowing, it turns out that the majority of them have been employed at least partially as conjunctions rather than as prepositions proper. This is in no way a contradiction in terms, since many of the most important Spanish prepositions, such e.g. para ‘for’, are frequently used as part of complex conjunctions in the donor-language, in this case: para que ‘so that’. I dare say that in the relative chronology of function-word borrowing the borrowing of conjunctional functions of such elements takes place prior to the borrowing of their prepositional functions.

Now, why is it that discourse particles and conjunctions preceed and outrank other grammemes in our borrowing scenarios? In a manner of speaking, these elements almost suggest themselves once it comes to hispanicizing your discourse. And discourse organization is something of a magic formula with regard to function-word borrowing from Spanish. Indeed, there can hardly be any doubt that, typologically speaking, Austronesian and Amerindian lan-guages are rather dissimilar in structure. The same holds true of the differences in status, numbers of speakers, duration of contact, degree of bilingualism, L1- and L2-literacy, and many more sociolinguistic parameters. The only thing that links all these languages together and that has triggered off processes which have eventually brought about a higher degree of surface similarity is their having been exposed to a dominant Spanish discourse tradition. If one wanted to be part of the colonial society or even wished to be successful in it, one had to adopt the discourse strategies which organized communication in the Spanish-speaking culture. Knowing how to argue was and is a prerequisite for participating in the distribution of power.

Imported and indigenous pragmatics of communication have not been identical. In post-conquest times, only the Spanish model defined the norms of how one was expected to connect the constituent parts of one’s discourse to one another. A short-cut to Hispanicizing one’s style of arguing appropiately seemed to call for the borrowing of exactly those Spanish elements which most saliently served the purpose of establishing discourse relations in an overt manner, viz. discourse particles and conjunctions like the four members of the core of cores which indicate adversative, alternative, causal, and anteriority relations, respectively.

However, the borrowing of function words does not automatically lead to a sudden over-all change in discourse organization. Instead of an unconditional Hispanicization of the textual principles, we very often find that the borrowed items do no more than add a Hispanic flavour to the way one expresses oneself. It is also a common feature of function-word borrowing that the resulting pairs of Spanish and indigenous synonyms are differentiated along the lines of stylistic parameters. Consider (32), for instance. The distribution of pero and its indigenous synonyms over formal and informal styles in Tagalog is representative of many such style differentiations in all three areas. Interestingly, it is most often the Hispanicized variety that is considered to be everyday style whereas the use of purely indigenous connectors or the like is most often associated with high-style language both in Austronesia and the Americas.

(32) Tagalog

informal pero ‘but’

formal kundi; datapwa't; nguni't; subali't ‘but’

Since the comparative investigation into the overseas contacts of Spanish has only just begun, there remains much more research to be done. Before I will set out to face such new tasks in future publications, I should like to conclude the present paper with a prediction (33). On the basis of what I have been arguing for in the above, one may assume that any new language added to the sample will follow the same patterns in function-word borrowing as those which have been established above, i.e. where there are more than two borrowed Spanish function words there will most probably also be a member of the core of cores - and this will most probably be pero.

(33) Prediction

For any given L_(MSP) iff it has borrowed function words from Spanish at all and iff the number of borrowed items exceeds 2, there is a maximum degree of probability that it has borrowed at least one member of C_(MSP) which is preferentially pero ‘but’.

Bibliography

Bills, Garland D.; Vallejo, Bernardo & Troike, Rudolph C.

1969 An Introduction to Spoken Bolivian Quechua. Austin, London: University of Texas Press.

De Pury-Toumi, Sybille

1979 "La logique de l’emprunt en mexicano." Amerindia 4

Gregores, Emma & Suárez, Jorge A.

1967 A Description of Colloquial Guaraní (= Janua Linguarum Series Practica 77). The Hague, Paris: Mouton & Co.

Hill, Jane H. & Hill, Kenneth C.

1986 Speaking Mexicano. Dynamics of Syncretic Language in Central Mexico. Tucson: The University of Arizona Press.

Mirikitani, Leatrice T.

1971 Speaking Kapampangan (= PALI Linguistic Texts: Philippines). Honolulu: University of Hawaii Press.

Porterie-Gutierrez, Liliane

1988 Étude linguistique de l’aymara septentrional (Pérou-Bolivie) (= Thèse Amerindia). Paris: A.E.A.

Preuss, Konrad Theodor & Ziehm, Elsa

1971 Nahua-Texte aus San Pedro Jícora in Durango. 2. Teil: Märchen und Schwänke (= Quellenwerke zur Alten Geschichte Amerikas aufgezeichnet in den Sprachen der Ein-geborenen 10). Berlin: Mann.

Reid, Aileen A.; Bishop, Ruth G.; Button, Ella M. & Longacre, Robert E.

1968 Totonac: From Clause to Discourse (= Summer Institute of Linguistics Publications in Linguistics and Related Fields 17). Norman: University of Oklahoma.

Schachter, Paul & Otanes, Fe T.

1972 Tagalog Reference Grammar. Berkeley, Los Angeles, London: University of Cali-fornia Press.

Smailus, Ortwin

1975 Textos mayas de Belice y Quintana Roo. Fuentes para una dialectología del maya yucateco (= Indiana, Beiheft 3). Berlin: Mann.

Stolz, Christel & Stolz, Thomas

1995 "Spanisch-amerindischer Sprachkontakt: Die ‘Hispanisierung’ mesoamerikanischer Komparationsstrukturen." Iberoamericana 58/59 (2-3), 5-42.

1996a "‘Transpazifische Entlehnungsisoglossen’. Hispanismen in den Funktionswort-inventaren beiderseits der Datumsgrenze." Areale, Kontakte, Dialekte - Dynamik der Sprache in mehrsprachigen Situationen. Beiträge zum 10. Essener Kolloquium vom 30.6.-1.7.1995 (= Bochum-Essener Beiträge zur Sprachwandelforschung 27). Werner Enninger (ed.). Bochum: Universitätsverlag Dr. N. Brockmeyer, 160-189.

1996b "Funktionswortentlehnung in Mesoamerika. Spanisch-amerindischer Sprachkontakt." Sprachtypologie und Universalienforschung 49 (1), 86-123.

1996c "Universelle Hispanismen? Von Manila über Lima bis Mexiko und zurück: Muster bei der Entlehnung spanischer Funktionswörter in die indigenen Sprachen Amerikas und Austronesiens." Orbis 39 (1). [in press]

Stolz, Thomas

1996 "Sprachkontakt, Sprachwandel und ‘Grammatikalisierung’ als Faktoren bei der Hispa-nisierung von indigenen Sprachen in hispanophonen Einflußzonen." Beiträge zur Teilsektion ‘Grammatikalisierung’ auf dem 24. Deutschen Romanistentag vom 25.-28.9.1995 in Münster (= Bochum-Essener Beiträge zur Sprachwandelforschung 29). Susanne Michaelis & Petra Thiele (eds.). Bochum: Universitätsverlag Dr. N. Brock-meyer. [in press]

Wolfenden, Elmer P.

1971 Hiligaynon Reference Grammar (= PALI Language Texts: Philippines). Honolulu: University of Hawaii Press.

Appendix 1: Abbreviations (used in the morpheme glosses)

Acc = accusative

Act = active

AF = agent focus

Apt = aptative

Ben = benefactive

Cau = causative

CFP = clause-final particle

Com = completive

Coo = coordinator

D2 = distal

D4 = unspecified deixis

Def = definite

Dur = durative

Emp = emphatic

Erg = ergative

Evi = evidential

Fin = finalis

Fut = future

Hum = human

Ina = inactive

Inf = infinitive

Lig = ligature

Loc = locative

NA = nomen agentis

Neg = negation

OF = object focus

Pl = plural

Prt = preterit

Psr = possessor

Rea = realis

Red = reduplication

Sg = singular

Top = topic

Trs = transitive

Appendix 2: Sample languages

1. AG Agta - Malayo-Polynesian, Northern Philippines - Philippines

2. AQ Ayacucho Quechua - Andean, Quechua - Peru

3. AY Aymara - Andean, Aymara - Peru/Bolivia

4. BT Batak - Malayo-Polynesian, Central Philippines - Philippines

5. BI Bikol - Malayo-Polynesian, Central Philippines - Philippines

6. CE Cebuano - Malayo-Polynesian, Central Philippines - Philippines

7. CM Chamorro - Malayo-Polynesian, isolated - Marianas/Guam

8. CH Chinantec - Oto-Manguean, Chinantecan - Mexico

9. CH’ Ch’ol - Mayan, Tzeltalan - Mexico

10. CO Cora - Uto-Aztecan, Corachol - Mexico

11. CQ Cochabamba Quechua - Andean, Quechua - Bolivia

12. GA Garífuna - Equatorial-Tucanoan, Arawakan - Honduras

13. GU Guaraní - Equatorial-Tucanoan, Tupi - Paraguay

14. HI Hiligaynon - Malayo-Polynesian, Central Philippines - Philippines

15. HQ Huánuco Quechua - Andean, Quechua - Bolivia

16. IN Inga - Andean, Quechua - Columbia

17. JA Jacaltec - Mayan, Kanjobalan - Guatemala

18. KA Kapampangan - Malayo-Polynesian, Northern Philippines - Philippines

19. MM Mam - Mayan, Mamean - Guatemala

20. MB Manobo - Malayo-Polynesian, Southern Philippines - Philippines

21. MK Mansaka - Malayo-Polynesian, Central Philippines - Philippines

22. MP Mapuche - Andean, Southern Andean - Chile

23. MZ Mazatec - Oto-Manguean, Popolocan - Mexico

24. MI Mixe - Mixe-Zoque, Zoque - Mexico

25. MX Mixtec - Oto-Manguean, Mixtecan - Mexico

26. NA Nahuatl - Uto-Aztecan, Aztecan - Mexico

27. NE Negrito - Malayo-Polynesian, Northern Philippines - Philippines

28. OT Otomí - Oto-Manguean, Otomían - Mexico

29. PG Pangasinan - Malayo-Polynesian, Northern Philippines - Philippines

30. PI Pipil - Uto-Aztecan, Aztecan - El Salvador

31. PO Popoloca - Oto-Manguean, Popolocan - Mexico

32. PP Popoluca - Mixe-Zoque, Zoque - Mexico

33. RP Rapanui - Malayo-Polynesian, Eastern Polynesian - Easter Island (Chile)

34. SB Sambal - Malayo-Polynesian, Northern Philippines - Philippines

35. TA Tagalog - Malayo-Polynesian, Central Philippines - Philippines

36. TG Tagbanwa - Malayo-Polynesian, Central Philippines - Philippines

37. TO Totonac - Totonac-Tepehua, Totonacan - Mexico

38. TZ Tzutujil - Mayan, Quichean - Guatemala

39. YA Yaqui - Uto-Aztecan, Taracahitic - Mexico/USA

40. YU Yucatec - Mayan, Yucatecan - Mexico

41. ZA Zapotec - Oto-Manguean, Chatino-Zapotecan - Mexico

[1] Earlier versions of this paper were presented at the universities of Rostock, Bremen, Kiel, Cologne, and Heidelberg, late in 1995 and early in 1996, respectively, and on occasion of the 1996 Winter Meeting of the Society for the Study of the Indigenous Languages of the Americas at San Diego (January 1996). I wouldlike to thank the public at all three places for their highly valuable comments. Likewise, I have to express my gratitude to my wife Christel Stolz who devoted much of her precious time to discussing the main issues of the paper with me. In order to save space, I have refrained from overloading the presentation with bibliographical references and the like. For more information (socio-historical, geographical, and historical) on Spanish-Amerindian and Spanish-Austronesian language contacts as well as for a next-to-complete bibliography of the pertinent literature, the reader is referred to our publications listed in the present bibliography.

[2] As for the choice of sample languages, I admit that it is a little biased for certain areas such as, e.g., the former Viceroyalty of Nueva Granada in the North of South America, are evidently underrated. It is hoped that future research will fill in these gaps as much as possible. For the time being, may it suffice to say that I expect to find - in those languages which have not yet been scrutinized - much the same phenomena as described here (cf. prediction (33) below).

[3] A remark on the character of my sources seems to be in order here. For obvious reasons, the bulk of my data is drawn upon second-hand information (grammars, dictionaries, text anthologies). Therefore, I am well aware of the inherent pitfalls of my sampling methodology. Some of my sources, indeed, may only be of a rather restricted reliability. However, for the present purpose, these sources may do for a temporary data-base, since all I want to prove this time is that there, in fact, are striking parallels in the borrowing behaviour of many different and unrelated languages which have been in contact with Spanish. This is also why there actually is no need for the sources to provide an absolutely complete inventory of grammatical hispanisms. In a number of cases, I guess, certain borrowed items do not show up in the grammatical descriptions although the ignored Spanish particles are well established in everyday language. In addition, note that I am concerned mostly with the iden-tification of borrowed grammemes rather than with the restructuring of syntactic patterns. For the point I want to make in the present paper, it is much more important to demonstrate that there is a principled convergence of Amerindian and Austronesian sets of borrowed grammemes. A complete catalogue of borrowings remains to be compiled in the near future.

[4] Talking of languages in this paper is a kind of short-cut way of dealing with the well-known problem of the dialectal heterogeneity of many of the Middle American, South American and Philippinian languages (or dialect clusters). The dialects which have been chosen for illustration in the present paper are not automatically viewed as the most authentic representatives of the dialect cluster to which they belong (by labeling the Durango dialect of Nahuatl simply as Nahuatl, for instance, it is only claimed that there is at least one identifiable dialect of the Nahuatl dialect cluster which displays the attested borrowing).