The purpose of this study is to demonstrate that variability with regard to the use of the so-called «personal a» can in some measure be explained by the importance of the role that marked nouns play in discourse¹²⁶. Role importance will be defined indirectly in terms of the semantic category of «individuation» as developed by Hopper and Thompson (1980). After describing the notion of «individuation», I will demonstrate through statistical sampling that it is a valid concept in the description of marked object nouns in Spanish at the phrase level. By referring to how the preposition is used with inanimate direct objects, I will show how the concept of individuation functions at the sentence level.

Finally, I will apply the notion to a case study from Hispanic literature. By demonstrating that the use of the preposition a with some object nouns can be associated with thematic literary importance, I will suggest that individuation also functions intersententially. As such, individuation provides a single, unified principle which explains the relationship of stylistics to meaning on the one hand, while maintaining a principle for categorical phrase and sentence-level uses of the preposition on the other.

I have limited this study to the examination of the preposition a as it occurs with direct object nouns denoting animals. Delimitation of this type allows me to focus on the greatest amount of variability in the preposition's use. As is well known, the use of the preposition with definite, human nouns is very frequent -a usage which no doubt underlies the misleading pedagogical terminology which labels the preposition the «personal» a. On the other side of the scale, inanimate nouns are most frequently, but not categorically, found without the marker. Animals, on the other hand, which are situated in the middle of an animacy scale, not only provide a rich source for the study of variability in language with regard to the preposition, but prove a challenge to those who seek an adequate description of its usage. The problem of adequate description is easily exemplified. The contexts found below, excerpted from a Latin American story, demonstrate the subtleties involved in the preposition's use with animals:

The fact that the dual mention of caballo refers to the same animal within the story line and that they are so close in proximity (in the same paragraph, in fact) seems to make them unlikely candidates for differentiation by the respective absence and presence of the direct object marker. Even more perplexing, perhaps, is that native informants, when asked about contexts such as these, see little difference in meaning for cases with or without the preposition. Despite the apparent flexibility in the use of the direct object marker, I would be reluctant to say that it transmitted no meaning whatsoever, for few linguistic tokens exist without carrying some functional load. The fact that such an unstressed particle could exist despite phonological tendencies to assimilate it to preceding or following vowels may attest to the strength of its semantic character.

Past scholarship on the use of a with animals is scant and generally limited to isolated comments in more general grammatical treatments. Many investigators stress the absolute uses of the preposition. Bello (1945: 283), among others, notes the categorical use of a with animals designated by a proper name. Isenberg (1968: 168) ties usage to the nature of the sentence's subject. Human subjects, according to this investigator, do not require an a with animal objects whereas animal subjects may. The Solé/Solé text (1977: 26) appeals

––––––––   224   ––––––––

to the ambiguity principle to explain the preposition's appearance with animals, i. e., the preposition must be inserted where, given freedom of word order in Spanish, the object might be incorrectly interpreted as the subject. Ramsey (1956: 46), on the other hand, underscores the variability of the preposition's use, noting its appearance when the speaker regards the animal as capable of reasoning or when there is a general desire to single it out from the surrounding context. Fish (1967: 80-83) develops this idea describing the use of a in general as «an affective reaction to something conspicuous». Conspicuousness is defined in terms of the status conferred on the animal or in terms of it being the object of a «precise» or unexpected action. He rejects more absolute treatments that associate the a with determinateness or the ambiguity principle.

In using terms such as «singled out» and «conspicuous», both Fish and Ramsey stress the marked noun's relevance to context and the optional, stylistic uses of the preposition. Since most of their examples are decontextualized, however, it is difficult to define the properties of a «conspicuous» noun. More importantly, we are not made to understand why such nouns should be singled out in the first place or what they have to do with other uses of the preposition.

I cannot presume in this study to resolve these complex issues in any absolute sense. Rather, with the help of a few recent theoretical notions concerning noun phrase roles in discourse, I hope to give the reader some insight into what a «singled o ut» noun phrase might look like, along with a glimpse of why and how such conspicuousness might be used.

During the last decade, focus in linguistics has shifted from studies underscoring sentence-level phenomena to those which bring context and language use to the fore. This, in turn, has occasioned a reworking of basic linguistic concepts. Take, for example, the simple concept of noun. At the sentence level, noun phrases are structural units fulfilling distinct functional and semantic roles (for example, subject as agent, direct object as patient, etc.). At the discourse level, the referents of these nouns are characterized as «participants» fulfilling roles in a series of states and events. Linguists have long recognized that not all participants play an equal role in discourse events. Some participants seem more salient relative to other participants in the same discourse.

One might use a theater metaphor to explain this concept. Suppose you were viewing a play. The stage, scenery, props, and actors are all participants within your conscious view. Yet some of these participants will demand more of your attention than others. You will focus primarily on the actors, whose motions will stand out against the inanimate props that serve as the stage setting. Some actors, namely the protagonists, will upstage other, relatively minor characters, unless of course, these minor participants momentarily seize center stage. As the story line unfolds there may be several changes in prominence, each of which will force a shift in the focus of the viewer's attention. Note that prominence might be related to characteristics inherent in the entity. For example, actors can move on their own volition while inanimate props cannot. Thus, by virtue of being an actor, a participant is more likely to be prominent than an inanimate prop. In other instances, prominence is related to elements outside the entity. The story line, for example, may bring a minor character to center stage for a moving soliloquy or a spot light may direct our attention to a brightly colored prop. Thus, prominence is the result of the complex interrelationship of an entity, its inherent characteristics and its contextual role.

Noun phrases in linguistic discourse play similar shifting roles. Much as the dramatist uses ploys such as role assignment, lighting, staging, or to assign salience, so the speaker employs grammatical devices. For example, assigning a noun phrase to be subject of the sentence, for reasons we shall see later, bestows greater discourse salience to the noun than other grammatical role assignments.

Direct object nouns, although less salient than subject nouns by virtue of their role, can have greater or lesser communicative importance. Nouns of relative importance have come to be known as «individuated» by virtue of the fact that they tend to represent separate, individual entities, and thus vie for our attention. More specifically, individuation, as defined by Hopper and Thompson (1980), is the degree to which object (patient) nouns are distinct from their subjects (agents) or from their background.

We might ask how it is that we are to recognize these nouns. In fact, there are characteristics,

––––––––   225   ––––––––

either inherent to the lexical noun or associated with the noun phrase, which tend to accompany individuation. Thus, animate nouns are more likely to be considered as separate entities than inanimate ones, specific nouns more so than nonspecific ones, singular ones more than those seen in mass, etc. According to this approach, object nouns are seen as appearing on a gradient where they are considered «more or less» individuated according to the types of properties that characterize them. One would say, then, that the referents of nouns reflecting those properties listed on the left-hand side of the scale given below are more highly individuated than the nouns with those properties given on the right: [Taken from Hopper and Thompson, 1980: 253].

HIGHLY INDIVIDUATED	LESS INDIVIDUATED
proper	common
human, animate	inanimate
concrete	abstract
singular	plural
count	mass
definite, referential	non-referential

Thus, individuation as a semantic category is not «defined» by one property but is rather «correlated» with varying clusters of properties. Nor is individuation an absolute, for nouns may differ greatly in individuation. In other words, individuation is a matter of degree, rather than a black and white category. More importantly for our study, such individuation has been associated with the syntactic marking of direct object nouns in several languages. It is obvious, in fact, that several of the characteristics that appear early in the individuation hierarchy (human, definite) are just those characteristics that have been used to explain the use of the preposition a with direct objects in traditional studies of Spanish. Such a hierarchy has not been applied, however, to lesser studied groups such as animals, nor has any correlation been sought with regard to variability of the preposition's use. In other words, we must ask ourselves whether we will find a greater incidence of the preposition with those animal direct objects which are most highly individuated. It is this task which I wish to address at this point.

In order to demonstrate that the factors associated with individuation were also found for the variable use of a with animals, I studied two contemporary novels which I chose purely on the grounds that they contained adequate numbers of animal direct objects. The one novel, Manuel Galvez's Pampa y su pasión (henceforth Pampa) deals with Buenos Aires's love of horse and track. The other novel, Miguel Delibes's Diario de un cazador (henceforth Diario), describes typical small game hunting scenes. Although both corpora contain comparably equal numbers of direct object animal nouns, there is an obvious difference in the way the animals are treated in the two works. The idea of honorific or elevated status granted to the animal by the assignment of a to the direct object, seems to be supported by the findings. Appeal to common sense might explain that the horse, described as noble and elegant in Pampa, should be marked more frequently than the small game animals that are typically chased, slain, and eaten in Diario. Indeed this is the case, for 72% of the nouns of Pampa are marked whereas only 18% of those of Diario carry the preposition. Furthermore, other attending linguistic features suggest statistically that the animals are treated in two distinct manners. As shown in the charts below, those of Pampa are more unique as indicated in their greater singularity, more identifiable as indicated by their definiteness, and more named as indicated by the number of proper nouns applied to their referents. Inasmuch as it is viable to accept naming, uniqueness, and identifiability as correlates of elevated status, we may accept the notion that the preposition a is used as an honorific with regard to animal direct objects. The chart below summarizes how the overall use of a correlates with the overall linguistic treatment of the animals as defined by selected properties on the individuation scale.

It should be pointed out that the significance of the scale does not have bearing on how animals may be marked in any absolute sense. It does not imply, for example, that

––––––––   226   ––––––––

horses are somehow deserving of more status than hunting dogs. Marking is relevant to such factors as importance in discourse, speaker intent, and relative speaker or cultural rating of the animal. Thus, it would be an easy task to find other corpora where horses are not marked at all. What the scale does show is that where the a is found, there are relatively high percentages of concomitant linguistic features that may be associated with status.

Those properties which relate to individuation may also be correlated more specifically with the markings of individual nouns. The following charts for each corpus summarize the presence or absence of a for specific noun types. As can be seen below, there is a scale with categorical or almost categorical usages at either extreme. Proper nouns, which by definition are uniquely identifiable, are categorically used with a. Indefinite plural nouns are at the other extreme with categorical absence of a in Diario and a 92% rate of absence in Pampa. In between the two extremes, percentages increase or decrease in scalar fashion. Although percentages differ across works, i. e., horses are marked more in all categories, the scale remains relative for each, i. e., within both works, proper nouns are more likely to be marked than definite singular nouns, definite singular nouns are more likely to be marked than definite plural nouns, etc.

Our pilot incursion into the uses of a with animals suggests a positive correlation between its appearance and the clustering principle of individuation: the more highly individuated the noun is (relative to other nouns in the same context) the more likely it is to be marked. It should be kept in mind, however, that the noun characteristics associated with the individuation hierarchy do not «cause» the noun prominence but are rather symptomatic of the role the noun plays within the speaker's message. Moreover, role importance is relative to that of other participants found in the same context -a context greater than the confines of the noun phrase. Thus we can safely speculate that there are other characteristics of individuated nouns that are not defined at the phrase level.

Since the definition of individuation involves the object noun's distinction from its subject and its background, it may be suggested that the properties associated with individuation at the noun phrase level are somehow related to the properties of other nouns in the discourse. Indeed, it turns out that the properties listed above to characterize individuated direct objects are more likely to characterize subject nouns. In other words, subjects are by general nature more likely to be separate individuals than objects. More to the point of this paper, when direct object nouns usurp the characteristics associated with subject nouns, they tend to be marked. In order to see the interrelationship between the object and subject nouns, the characteristics of these roles must be discussed.

Givón (1976, 1979) has shown that subject nouns are overwhelmingly definite and often human. Due to communicative prejudices, we tend to talk about human beings rather than other categories of noun referents. Thus, they tend to be the topicalized center of our speech¹²⁷. So much is this the case that in some languages, there are restrictions on the indefiniteness of subjects. Languages such as Spanish and English that allow indefinite subjects do so on a very limited basis. Object nouns, on the other hand, are at the other end of the spectrum. They are often inanimate and indefinite. In fact, one of the major tasks of the object noun is to introduce new information into the discourse, thus explaining why they show up as indefinite more than any other noun phrase. When object nouns do take on properties similar to those of the subject, it may be argued that they compete for the communicative spotlight and thereby, may require marking.

The noun phrase characteristics associated with individuation correlate with the nouns'

––––––––   227   ––––––––

potential activity level at the sentence level, i. e., highly individuated nouns tend to be more «active» than non-individuated nouns. As pointed out by Garcia (1975), grammatical subjects, many of which are agents, tend to be, relative to other nouns, the most active participants of the sentence¹²⁸. Direct objects are at the other end of the spectrum as the least active or most totally manipulated. Indirect objects are somewhere in between. Support for these notions can be found in the similarity of indirect objects and agent subjects: Both tend to be definite and animate. It is not fortuitous, then, that the indirect object, in order to distinguish it from the subject, must be marked, not coincidentally, by the way, with the very same preposition that marks highly individuated direct objects. One would expect that this would lead to a very gray, indeterminate area between direct and indirect objects, which is exactly what we find in pairs such as Le ayudo «I help her» (indirect)/ La ayudo «I help her» (direct)¹²⁹.

The diagram given below graphically represents our preceding discussion. Subjects are separated from competing objects completely by markers. The difference between indirect and direct objects is not clearly marked but characterized by indeterminacy. On either extreme of the spectrum, subjects and the least active, least individuated direct objects are not marked at all. They are too far apart semantically for it to be necessary.

Using this theory, I demonstrated (Weissenrieder, 1985) that inanimates are obligatorily marked with verbs of placement in instances where the lexical verb takes objects that have the same activity level as their subjects. Thus, in sentences 1a and 1b, the lack of agentness on the subject's part enables the equally inactive object to be competitive with the subject role. Marking is therefore necessary:

Contrastively, where the direct object plays a typically inactive role overshadowed by an agent subject, marking is not permitted. This is demonstrated in the pair of sentences below where the professor is an agent in 2a and a manipulated pawn in 2b. Sentence 2b is similar to 1a and 1b and marking is obligatory.

Competing activity level remains a valid concept in explaining variable uses of the preposition with animals. Consider, for example the scene given below, taken from Diario, where a priest rather comically describes to a character named Pepe what heaven is like for hunters. I have italicized and numbered the animal direct objects for reader convenience.

Despite eight mentions of animals as direct objects, only number 6 carries the marker. The sentence in which it is found is distinctive in that subject and object contrast mutually. Linguistically, the contrast is evidenced by the presence, in the clause that follows the preposition, of the emphatic expression si que, the repetition of complete noun phrase subject las perdices, and the contrastive tag phrase a ti¹³⁰. The verb ver plays an important role since it allows semantic reversibility of the subject and object roles, i. e., both humans and partridge are equally capable of seeing¹³¹. More importantly, who sees whom is a crucial question in the discourse. Note that this is not to say that subject and object cannot be distinguished in their grammatical roles. There are at least three obvious cues that signal tú as the subject of the sentence (the person and

––––––––   228   ––––––––

number marking of the verb veas; the position of tú before the verb; the case marking of tú, rather than te or ti, etc.). Thus, there is no role ambiguity in the sentence, but rather a commentary on the potential equivalence of activity level on the part of the two nouns. They are made, so to speak, to compete for center stage¹³².

High activity roles such as subject are, as has already been pointed out, associated with high topicality. Since they are the participants spoken about most, they tend to draw more of the communicative focus. When less active roles such as direct object are assigned greater thematic importance, we can expect them to be singled out and marked.

We now turn to a Latin American short story, where the author has manipulated the use of the a for these reasons. Since the preposition's appearance is linked to discourse notions, it's important to have an idea of the main characters and their relationship.

There are three participants in the story. Gustavo, an adolescent who finds himself alone in an amusement park, a stray cat that Gustavo picks up and befriends while in the park, and a girl Gustavo's age whom he meets the same night, subsequent to his adoption of the cat.

The little girl is described as soft, perfumed, and timid. She is attracted to Gustavo but dislikes the cat. Gustavo is characterized as lonely and is attracted to both the cat and the little girl. The triangle that results from both the cat and the girl vying for Gustavo's attention creates an interesting milieu for variability with regard to direct object marking.

Linguistically, Gustavo, the protagonist, controls the thematic spotlight. As predicted, this correlates with an overwhelming dominance of the subject role. He represents 42% of all referents functioning as subject. The cat follows in thematic dominance with 18% of the subject role; the little girl is a close third with 11%. Despite the relatively important thematic focus given the cat, it is never linguistically personified. In other words, it is never the subject of any lexical verb not normally associated with animals, nor are any selectional restrictions on verbs violated when it functions as direct object. In fact, the verbs with which it appears are neutral with regard to the animacy of their objects¹³³.

The cat appears in the direct object role in nine instances. Although it is highly individuated in all cases -at least in terms of noun phrase characteristics, only five cases are marked. As previously suggested, this indicates that some of the correlates of individuation are found in diffuse ways in discourse, and therefore, are not as easily countable as the correlated properties of noun phrases. Nonetheless, it can be shown that the preposition appears when the cat is found in thematic focus and is absent when the young lady takes over center stage.

Consider, for example, the contexts below. When, in the initial stages of the story, Gustavo and the cat are the only participants, the cat has Gustavo's complete attention and is marked. When the subject ellos is implied, it can be easily interpreted as cat and boy. The cat is further singled out by the use of topicalized word order and the surrounding discourse which deals clearly with the cat as topic:

When, on the other hand, the little girl becomes dominant, the cat's participation is minimized and no marking appears. Consider the following context, where the mention of the cat is found in standard SVO/Adverbial word order and is surrounded by discourse which is clearly dedicated to the description of the little girl:

Nor does marking appear in the following two contexts, where Gustavo and the little girl are the more focused participants. Focusing is indicated by the inclusion of the girl in the implied ellos and the dialogue with her after the mention of the cat. Again, word order is

––––––––   229   ––––––––

standard:

By contrast, the a is used when the cat is in direct parallel with the girl. It is interesting to note that in the second context, the little girl seems to be winning Gustavo's attention as is indicated in discourse by the description of the metaphorical fusion of their bodies and the corresponding reaction on Gustavo's part. Interestingly enough, the preposition marks both the cat and the inanimate hand of the little girl. Word order is again marked:

As this story demonstrates, a direct object noun can be singled out of context based on its thematic level of activity relative to other noun participants. The a of this story becomes a discretionary, stylistic marker that the author has carefully used to harmonize with the interplay of the three characters. No matter what symbolic role we might assign him, the cat represents a literary pawn whose evershifting level of dominance is linguistically signaled¹³⁵.

It should be pointed out that not in every corpus where the a is variable with animals is discourse as clear a factor as it is in this one. Elevated status can be assigned through any number of channels. In some instances, verb phrase characteristics personify animals at the sentence level and consequently, marking seems to be a reflex of syntax rather than a variable influenced by discourse. At the other extreme, marking may be linked so diffusely with some symbolic role or greater thematic value that any explanation of its usage can only be substantiated at a highly inferential, interpretive level. Worse yet, both tendencies can be found within the same discourse. In short, investigation of the variable uses of the preposition is painfully slippery and often not very productive.

In the short story examined here, however, verb phrase and noun phrase characteristics were constant, allowing us to examine the ways in which discourse alone might influence the presence of the marker. More importantly, the discourse itself provided other concomitant features which freed us from relying too heavily on personal interpretation in determining the thematic importance of the participants. To this investigator's delight, it also offered a glimpse of how, as theorized, the principle of individuation could be expanded to include variable uses of the preposition with inanimates.

In this study, I have suggested that individuation, as defined as a noun's distinction from its subject and its background, is a valid concept with regard to the marking of direct object nouns denoting animals. Individuation derives its motivation from the varying roles noun phrases play in discourse. Noun phrase properties, (animacy, definiteness, etc.) may be used as general heuristics in distinguishing highly individuated objects at the phrase level. At the sentence level, such characteristics are typically associated with the active, topicalized role of subject which in turn, must be separated from similar potential for individuation in the indirect object and in some direct objects by the marking of the latter two. At the discourse level, subject-like properties of thematic importance may be bestowed on the direct object and thus also motivate marking.

More specifically, at the phrase level we demonstrated that there is a gross statistical correlation between individuation and the variable use of a. There are obvious limitations to any study which focuses on noun phrase characteristics alone. There is no one-to-one correlation between individuation, as it is identified by noun phrase properties, and the appearance of the a, i. e., highly individuated animal nouns are not necessarily marked, but where marking does occur, highly individuated nouns are more marked than less individuated ones. This lack of correlation occurs because noun phrase characteristics do not define individuation but are rather symptomatic of it.

At the sentence level, the concept of individuation was broadened to include activity level, i. e., highly individuated nouns tend towards high activity levels. Thus, activity level correlates with noun phrase characteristics

––––––––   230   ––––––––

e. g., human nouns are highly active) as well as with the noun's syntactic role (e. g., definite, human nouns tend to be the subject). Marking occurs when the direct object competes in activity level with the subject.

At the discourse level, competition between nouns is often intersentential. By using a Latin American story where the interplay between three characters played a major role in literary theme, I have suggested further elaboration of the principle of individuation. Potential for individuation in this story was identified not through the properties associated with noun phrases, nor through sentential activity level, which were consistent throughout the story, but by contextual factors such as word order, thematic role, and participant identity in the discourse immediately preceding and following the direct object noun in question.

Both a strength and a weakness in the concept of individuation is its potential for varied application. Recognition of the fact that direct objects may be singled out of a context does not in any way predict how the strategy will be put to use. In this study, we have seen the a as marker of status, contrast, and thematic importance of the direct object noun. Given the limitations of this study, other strategies for use, as well as how these meanings might be derived have been ignored.

The value of individuation lies in that it does provide a single concept which underlies distinct uses, whether they be at phrase, sentence, or discourse level. Thus, it enables us to explain why a might be used with both animate and inanimate nouns. Moreover, it provides insight into variability in language by linking the black-and-white choice of the preposition's selection with the gray indeterminacy of a range of meaning. The logical result is the unpredictability in the use of the preposition with direct objects such as animals, the inability of native informants to describe differences in meaning for cases with or without the a, and the subtle manipulation of the preposition's use for personal and stylistic effect.

Bello, Andrés y Rufino J. Cuervo. 1945. Gramática de la lengua castellana. Buenos Aires: Sopena.

Bernardo, Robert. 1980. «Subjecthood and Consciousness» in William Chafe, ed. The Pear Stories 275-99.

Chafe, William L. 1980. The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production. Norwood, NJ: Ablex.

Delibes, Miguel. 1966. Diario de un cazador. Barcelona: Destino.

Fish, Gordon T 1967. «A with Spanish Direct Object» Hispania 50: 80-85.

Gálvez, Manuel. 1958. La Pampa y su Pasión. Buenos Aires: Losada.

García, Erica. 1975. The Role of Theory in Linguistic Analysis: The Spanish Pronoun System. Amsterdam: Elsevier North Holland.

Givón, Talmy. 1976. «Topic, Pronoun and Grammatical Agreement». In Charles Li, ed. Subject and Topic. New York: Academic Press 149-88.

___. 1979. On Understanding Grammar. New York: Academic Press.

Guido, Beatriz. 1967. «El niño en el arco» in Rodríguez-Alcalá, ed. Cuentos nuevos del sur. Englewood Cliffs, NJ: Prentice Hall 33-39.

Hopper, Paul J. and Sandra A. Thompson. 1980. «Transitivity in Grammar and Discourse». Language 56: 251-99.

Isenberg, Horst. 1968. Das Direkte Objekt im Spanischen. Berlin: Akademie Verlag.

Quiroga, Horacio. 1968. Cuentos. México: Porrúa.

Ramsey, M. 1956. A Textbook of Modern Spanish. New York: Holt.

Solé, Yolanda and Carlos Solé. 1977. Modern Spanish Syntax. Lexington, MA: Heath.

Weissenrieder, Maureen. 1985. «Exceptional Uses of the Accusative a». Hispania 68: 393-98.

Hispania [Publicaciones periódicas]. Volume 73, Number 1, March 1990