The Castilian reflexes of ABHORRERE/ABHORRESCERE: a case-study in valency.
 

Appeared in Hommages offerts à Maria Manoliu-Manea, ed. Coman Lupu and Glanville Price (Bucharest: Pluralia/Logos), 1994, pp.122-48.
 

Christopher J. Pountain, Queens' College, Cambridge, United Kingdom
 
 

1 Introduction

The purpose of this paper is to study the factors involved in the changing valency of the reflexes of ABHORRERE/ABHORRESCERE in Castilian, ie aborrir~aburrir and aborre(s)cer. I use the term 'valency' in a more restricted sense than is sometimes usual[1] to refer especially to properties of 'voice' and 'transitivity', although these properties in their turn relate to a number of combinatorial features of verbs. My main contention will be that while the semantic history of aborrir~aburrir in particular may be understood in simple terms as comprising a change in valency, the factors which bring about that change are extremely complex, involving a detailed consideration of the position of these verbs within the morphological, syntactic and semantic structures of Castilian.
 
 

2 Preliminaries

1.1 Some basic notions of valency

I propose to adopt the two traditional basic valency categories of transitive and intransitive in the classification of verbs. Within transitive I will make two further distinctions: active, mediopassive and passive and causative and non-causative.

I shall include in the category of transitive any verb which has more than one argument: this is a broader notion of transitivity than the traditional one, but one which is in tune with the approach of modern linguistic theory[2].

Passivisation is traditionally taken to involve any syntactic device through which an object of a transitive verb comes to function as its syntactic subject: put more generally, we may say that it is a particular case of arguments of a verb changing their surface syntactic roles. I use the term mediopassive primarily for passives in which an agent is not directly involved and so is not normally expressible. However, the boundary between mediopassive and passive is hazy in Castilian, since the reflexive, which is the archetypical exponent of the mediopassive, is increasingly also an exponent of the passive[3]. In such cases I shall hedge my bets by using the term 'mediopassive?'.

In this article I shall speak particularly of active and passive valency in past participles. The valency of the past participle of a transitive verb (both non-causative and causative) is normally passive: thus non-causative visto = que ha sido visto, causative enfurecido = que está enfurecido. The valency of the past participle of an intransitive verb is normally active: thus salido = que ha salido. However, some past participles have unexpected valencies: Castilian cansado, for example, may be said to have both passive and active valency in that it can mean both 'tired' (= que está cansado) and 'tiring' (= que cansa). The past participle of a (mediopassive?) reflexive verb is passive with respect to the corresponding non-reflexive verb and active with respect to the reflexive verb itself: thus enfurecido is also active with respect to enfurecerse.

All causative verbs are transitive, since the notion of causation necessarily involves more than one verbal argument, but transitive verbs can also be non-causative (Castilian ver 'to see', is an example of a non-causative transitive verb). The only causative/non-causative relation that is to any extent marked by morphology in the Romance languages is that between a causative and a corresponding intransitive or mediopassive non-causative, eg Castilian cerrar 'to shut', ie 'to cause to shut (intransitive)', cerrarse 'to shut (intransitive)' or 'to get shut (mediopassive)'. A special case of this relation is that which holds between causatives and a corresponding mediopassive? reflexive verb which expresses a mental state, eg asustar 'to frighten' / asustarse (de) 'to be frightened (by)'. The difference is that verbs like asustarse may take a (prepositional) object (which has in fact the function of an agentive) whereas verbs like cerrarse on the whole (though the situation may be changing[4]) do not. In such a case, the object of the causative will be the subject of the non-causative, and the relation between pairs like cerrar and cerrarse therefore resembles that between active and passive. The transitive causative/non-causative relation is usually expressed lexically, eg ver 'to see', hacer ver 'to cause to see'. Here the subject of hacer is different from that of ver. The object of ver in hacer ver is coreferential with the subject of neither hacer nor ver, although such coreferentiality may be expressed by the use of a reflexive (hacerse ver). In order not to lose sight of this important distinction, I will subdivide causatives into causative(intransitive) (eg cerrar), causative(mediopassive?) (eg asustar), causative(transitive) (eg hacer ver), and causative(reflexive) (eg hacerse ver)[5].

In the case of ABHORRERE/ABHORRESCERE, we will be considering two valency changes: (a) the movement of medieval aborrir 'to hate' to modern aburrir 'to bore', which may be thought of as a transitive non-causative (X hate Y) -- causative(reflexive) (Y cause [X hate Y]) shift, and (b) the extension of aburrido from the meaning of 'bored' to that of 'boring', ie a passive -- active shift[6].
 
 

1.2 The semantic field of 'antipathy'

The semantic field to which ABHORRERE/ABHORRESCERE belongs may be broadly labelled as that of 'antipathy'. It is probably not desirable to begin by setting up any more specific notions than this as semantic primitives, since this field clearly involves gradient notions which are differently discriminated from language to language. I therefore propose the general transitive verbal notion of

(1) (X) HAVE ANTIPATHY FOR (Y)

and its causative(reflexive) counterpart

(2) (Y) CAUSE [(X) HAVE ANTIPATHY FOR (Y)]

as two fairly basic semantic notions which will facilitate our discussion. In English, hate would fall towards the 'strong' end of the scale of (1) (loathe would be right at the top of this scale and dislike would be near the bottom); bore would fall towards the 'weak' end of the scale of (2) (annoy being 'stronger' and disappoint being 'weaker').

In tracing the semantic trajectory of ABHORRERE/ABHORRESCERE we shall find that we have to consider not only this semantic area of human-related emotion or mental attitude, but also a more abstract relation of incompatibility between things. We shall also have to consider action that is consequent upon antipathy and incompatibility (renunciation, rejection, etc). These notions can be related diagrammatically as follows:

(3)

Changes in meaning in this area observable within the Romance languages excluding Ibero-Romance appear to involve shifts in semantic strength within each of categories (1) and (2), but not shifts between categories (1) and (2). Changes in semantic strength in this field are not of course a priori surprising: in fields of meaning which relate to human emotions it is common to encounter the force of hyperbole leading to an overall tendency to attenuate meaning[7]. Within Latin itself we find ABHORRERE itself ranging from the extremely strong 'etymological' meaning 'to shrink back from' to the relatively weak sense of 'not to agree with'[8]. French ennuyé is in origin the past participle of a verb *INODIARE, presumably a causative counterpart to *ODIARE (based on ODIUM 'hatred'[9]); hence *INODIATUS[10] 'caused to hate' has attenuated to '(caused to be) bored', and predictably we find in Old French the stronger intermediate value of enoiier as 'to be burdensome to' (also approximately with this meaning in the borrowing into English as annoy)[11].

Indeed, outside Ibero-Romance categories (1) and (2) have remained well discriminated. In Italian, for example, despite the original morphological relation between Latin *INODIARE and ODIARE, there is now no obvious morphological relation between odiare 'hate' and (an)noiare 'to bore, annoy': odiare is a learned word, a denominal formation from Medieval Latin odium; (an)noiare is usually[12] thought to be a borrowing from Provençal enojar. In Old French, the borrowing haïr from Frankish *HATJAN rendered the idea of 'hate'; in Romanian, where *ABHORRIRE 'hate' survived as a urî[13], 'bore' is rendered by a plictisi (<Greek pleektoo).
 
 

1.3 The position of Ibero-Romance

Two factors appear to set the Ibero-Romance languages apart from general Romance. The first is the strong survival of ABHORRERE, which was not generally favoured (though in this, as in many other lexical features, Ibero-Romance and Romanian do have common ground). The second is the predilection of the Ibero-Romance languages for verb-forms which have been 'strengthened' by the inchoative suffix -ESCERE[14], in the light of which the appearance of a form deriving from ABHORRESCERE[15] is not especially remarkable. Nor, probably, is the assumption of a period of competition (for Portuguese and Castilian) between an inchoative verb and its non-inchoative counterpart with similar meanings: several such pairs of verbs are attested. Some examples from Castilian are fallecer/fallir 'to fail', bullecer/bullir ' to boil', guarecer/guarir 'to protect, shelter', podrecer/pudrir 'to rot', florecer/florir 'to flower' (the surviving form is in each case underlined)[16]. Sometimes a fossilised past participle remains as an adjective in the language, thus fallido 'vain, unsuccessful', florido 'flowery', and (as a noun) guarida 'den, lair'. What is more unusual, however, is to find a long continuing period of coexistence, since either one form or the other usually becomes obsolete; I have been able to find very few surviving doublets parallel to aburrir and aborrecer[17]. It would be quite natural, therefore, to find a reflex of ABHORRESCERE steadily growing in popularity in Portuguese and Castilian and eventually ousting the reflex of ABHORRERE in the general meaning of 'to hate'.
 
 

2 Castilian

2.1 The coexistence of aborrir and aborrescer in Old Castilian

The following comparative table (4) is built up according to the definitions (slightly modified to show transitive valency more explicitly) given for each word, and the classification of those definitions, given in Müller, 160-6, 175-7, and plotted as for (3):
 

(4)

To this table we must add for aborrir the meaning (5.5) 'abolir [algo]', which may be a contamination from abolir itself, and for aborrecer the interesting example (5) with the meaning given as 'inspirar repugnancia', which would appear to be an early causative value for this verb (and the meaning eventually adopted in attenuated form by aburrir).

It would seem at first sight that aborrir and aborrescer coexisted in Old Castilian with a range of meanings which overlapped to a considerable degree. However, closer inspection of the very wide range of examples given in this invaluable source of reference is instructive. It must be remembered that the data presented derives from a great number of authors, periods and styles, and we must therefore be prepared for the possibility of individual preference within these: Berceo, for example, seems not to favour aborrescer.

Even the most cursory inspecton of the data in (4) would suggest the following distinctive properties of aborrecer:

From Müller we can also see that

As regards aborrir, it is also perhaps significant that a great many examples given under aborrir are actually past participle forms (aborrido is dealt with separately by Müller only in its 'adjectival' meaning), and that aborrir appears never to have been used as a translation of the Vulgate ODI (and indeed appears to be absent from Biblical translations such as the Biblia Romanceada).

Other derivatives show expected valencies. The various nominalisations (abor(r)imiento, aborrencia, aborricion; aborrencia, aborrecimiento) all carry the meaning of 'hatred', or 'thing that is hateful'; the disappearance of all but the two forms in -miento can no doubt be attributed, as Müller suggests[18], to a lack of significant semantic difference among them (and aburrimiento and aborrecimiento were to be discriminated in Modern Castilian).
 
 

2.2 The Sixteenth Century onwards

Cuervo, 80, reports that aburrir conserved its medieval meaning ('to hate'), in the literary language of the 16th Century and in popular language in the 17th. The last example chronologically that he cites is from Jovellanos (18th Century). The meaning is marked as antiquated in the 1770 Academy Dictionary (DHRAE, 249). The new valency of causative(mediopassive?) alongside the old transitive non-causative valency appears to be attested in the Oudin dictionary of 1607, where aburrir is glossed as 'fascher [causative(mediopassive?)], hayr, desesperer [causative(mediopassive?)], auoir horreur, se desgouster'. (We can probably take 'se desgouster' as representing the reflexive usage; I have in fact been unable to find a clear example of non-reflexive aburrir in causative(mediopassive?) valency before the DHRAE's example, 250, 'A oírme, si no os aburro' from Moreto (17th Century). Neither Cuervo nor DHRAE record an intransitive valency for non-reflexive aburrir[19].) These new values of aburrir occupy a different area of the field of 'antipathy' from the present-day meaning of 'to bore, annoy'. The idea of 'to cause to be dispirited, desperate' and even more for the past participle aburrido the idea of 'unhappy, in despair', is both a stronger meaning and one which has to do with the ideas of incompatibility and rejection plotted in (3) (see (15)).
 
 

2.2.1 The development of the reflexive aburrirse

Aburrir also develops a reflexive construction which dates from the Golden Age: Cuervo gives examples from Lope, Cervantes and Tirso. Aburrir never had a reflexive construction at all, so far as we can tell, in the medieval language (though aborrecer did, cf. ((4), meanings 3.2, 6 and 7)). Corominas & Pascual (I,25) see the development of the reflexive construction as the necessary step in the development of the new valency in non-reflexive usage. Quite how this might have come about, however, poses interesting questions. We may identify two important categories of reflexive in Castilian for the purposes of discussion: (a) 'literal', 'figurative' or 'weak' reflexives which maintain the valency of the corresponding non-reflexive verb (eg lavarse 'to wash oneself', encontrarse 'to find oneself', hence 'to be (in a place or situation)', llamarse 'to call oneself', hence 'to be called'), and (b) 'intensifying' reflexives which effectively reverse the valency of the corresponding non-reflexive verb (eg olvidarse de 'to forget', but also olvidar with the same meaning). Now if an aburrirse form of category (a) had developed (ie in the meaning of 'to hate oneself'), it would not have challenged the valency of aburrir. And if the new aburrirse was an 'intensifying' form of aburrir (let us suppose *aburrirse (de) with the meaning of 'to hate'), it would still not necessarily have challenged the valency of aburrir. We can only make the link between reflexiveness and valency change if we assume that a reflexive of one kind was subsequently re-interpreted as the other kind. The process, represented graphically in (6), must be construed in roughly the following terms. Let us imagine that aburrir 'to hate' (A) developed the intensifying reflexive form aburrirse (de), also meaning 'to hate' (B), and that aburrirse (de) is then construed (D) to be a verb of mediopassive? valency like horrorizarse (de) 'to be horrified (by)' (F). At this point we would have to assume that aburrirse (de) had extended its meaning from 'to hate' to 'to be annoyed by' (D). The way would now be clear for the creation of an analogical aburrir (C1) as a causative(mediopassive?) in the meaning of 'cause to be annoyed' in the same way as the causative(mediopassive?) horrorizar means 'to cause to be horrified' (E). The new causative(mediopassive?) aburrir (C1) is semantically close to the notion of 'to cause to hate' (C2), ie the causative(reflexive) counterpart of the 'old' aburrir (A)[20].

(6)
 
 

Cuervo, 80, similarly sees the development of the reflexive aburrirse as consistent with the cases of semantically related verbs of the same valency:

although he is here assuming the new causative(mediopassive?) valency for aburrir. But Corominas & Pascual, 25, convincingly place the reflexive usage chronologically ahead of the non-reflexive usage. Under the heading of the new valency, Cuervo, 80, and DHRAE, 250, indeed quote reflexive and past participle examples from Golden Age authors but hardly any non-reflexive examples. Even Autoridades illustrates aburrir exclusively with reflexives.

The role of the past participle aburrido in the change of valency seems to me to warrant further investigation and comment. We have already noticed how very frequently the past participle figures in the medieval and Golden Age examples cited in reference works; Covarrubias, 30, whilst lumping aborrecer and aburrir together (see below) with almost identical definitions, specifically defines aburrido as 'el que de sí mesmo está descontento, despechado y determinado a perderse, sin reparar en el daño que se le puede seguir' as opposed to aborrecido 'el desechado y mal visto', which might suggest that aburrido had a certain independence from its corresponding verb, as indeed do many past participles which are frequently used in an attributive adjectival sense. It would not be impossible, then, to visualize aburrido as evolving independently of aburrir. Müller, 173, in fact suggests tentatively that both aborrido and aborrescido showed signs of extending their meaning in the medieval language from the literal passive past participle meaning 'that which is hated' to a meaning more generally associated with hatred: 'hateful' in the case of aborrescido and (possibly) 'having hatred' in the case of aborrido; examples for the latter are worth quoting in full:

(7)

We must also bear in mind that pragmatically speaking an object of hatred may also be (and indeed in most cases probably is) responsible for inspiring that hatred) (eg 'the hated totalitarian régime'). It would therefore be relatively easy for the past participle aburrido to extend its meaning from 'that which is hated' to 'that which causes or inspires hatred', ie from passive to causative(reflexive) valency. But while such a value may have been an intermediate stage in the evolution of aburrido, it is not the new value chiefly described in the 17th Century: here, as we have seen, the meaning is 'dispirited, desperate'. I refer back to the meanings of aburrir plotted in (4) under the headings of 'abstract relation' and 'action'. Meaning 5.1 of (4) 'to reject, despise' probably provides us with the link: a person who feels rejected and despised (by others or himself) has just the quality of dispiritedness or desperateness implied in 17th Century aburrido. DHRAE ,248, also points to the importance of this area of meaning of aborrir~aburrir.

A second factor to be borne in mind is the potential denominal nature of the past participle inflection. There are a number of such exclusively denominal 'past participles' in Castilian (eg alado 'having wings (alas)', which has no corresponding verb *alar). It is certainly not impossible to envisage aburrido as having developed a meaning of 'associated with aburrimiento', ie 'hateful' rather than 'hated', in this way, and hence having widened its range from that of acting exclusively as the deverbal past participle of aburrir.

If indeed aburrido evolved somewhat independently in either or both of these ways as an adjectival past participle (see also the remarks on aborrido below), the application of the standard pattern of valency relations (aburrir (causative(mediopassive?)), aburrido (passive past participle)) might well have encouraged, and at the very least would have been entirely consistent with, the development of a reflexive aburrirse (non-causative mediopassive?) in the sense of (following Covarrubias, 30) 'ponerse descontento, despechado y determinado a perderse'.
 
 

2.2.2 Further differentiation between aburrir and aborrecer

I now return to the 'contest' between aburrir and aborrecer. The matter is clearly more complex than the lists of examples given in Cuervo and elsewhere suggest. We may confidently suppose that aburrir was a good deal less popular than aborrecer in the 16th and early 17th Centuries. Nebrija does not mention it or any of its derivatives in his Vocabulario de romance en latín (1516), although he lists aborrecer, aborrecible, and aborrecedor. Neither does De las Casas (1576). Although aborrir, aburrir and aborrecer are glossed in the Oudin dictionary (1607, 1645, 1660), only aborrecer is mentioned in any glosses of French words in this field (as the first equivalent for haïr), which shows at least that aburrir did not suggest itself readily. Even in a work as late as Sobrino (1705), aburrir and its derivatives do not appear in the glosses for French words such as fâché, ennuié, désolé, etc., although they are given in the Castilian-French section of the work. Aborrecer and aburrir also appear to have belonged to different registers: Covarrubias, 30, gives a very clear sociolinguistic index to aburrir, describing it under his entry for aborrecer as a 'término más grosero' with the same meaning. DHRAE, 248, notes the popularity of aburrir in the language of Juan del Encina's imitators of rustic speech.

Nevertheless, evidence from 17th- and 18th-Century dictionaries shows aburrir steadily pulling away from aborrecer in meaning. Oudin (1607) shows a certain semantic overlap between the two in the sense of 'hate', though aburrir is additionally glossed as 'se desgouster'; derivatives of the two verbs, however, contrast quite markedly:

(8)

In Autoridades (1726), aburrir is clearly distinguished as distinct in meaning from aborrecer and is clearly causative. Aburrir is glossed as 'apesadumbrar mucho, hacer despechar y desassossegar à uno, de suerte que no solo le entristezca, sino que casi llegue a aborrecerse'.

Nevertheless, aborrecer also showed signs of developing the causative valency eventually taken on by aburrir. Several 16th- and early 17th-Century examples are quoted by Cuervo. Aborrecer is also recorded as having developed a reflexive form similar in meaning to the new aburrirse. This in fact is the direction that aborrecer was to take in Portuguese, eventually ousting aborrir altogether from the language.
 
 

2.2.3 Aborrir and aborrido

It is easy to assume that aburrir simply replaced aborrir by the analogical generalization of what were originally radical-changing variants in the paradigm (aburrió, etc.)[21] Nevertheless, it is striking how many sources list aborrir and aborrido as being distinct from aburrir and aburrido. Oudin's entry is as follows:

(9)

Minsheu (1623) has a separate entry for aborrir which refers the reader to aborrecer, while aburrir is given with a completely different meaning ('to despair, to be forlorne, to be past all hope'). De la Porte (1659) also gives more or less identical glosses for aborrescer and aborrir ('haten (to hate)/aftreck hebben (to despise)'), but glosses aburrir significantly differently glossed as 'hem stooren (to inconvenience someone)/haten (to hate)/wanhopen (to despair)/schroom hebben (to be fearful)'). Again, however, there is considerable doubt as to whether the semantic trajectory of aborrido followed that of aborrir. Aborrido is less well documented; but where it is, it seems to be regarded as an alternative to aburrido rather than to aborrecido. We have already noted Oudin's gloss (9), which corresponds much more closely to that given for aburrido than for aborrescido (8); and Autoridades, 17, finally sounds the death-knell of aborrido in the following terms: 'Es voz antiquada, porque ya se dice aburrido'. Thus while in its relatively short post-medieval history aborrir retained something of its medieval meaning, aborrido appears rather to have been considered a variant of aburrido.
 
 

2.2.4 The dual valency of aburrido

In modern Castilian, aburrido has come to have dual valency, adding to its range the meaning of 'boring'.

Semantic analogy may have played an important role. Today aburrido is a member of a small class of semantically-related verbs whose past participles have a double valency, its closest semantic relations being cansado 'tired / tiring' and the antonymous divertido and entretenido 'entertained / entertaining'. Another close relation is pesado 'weighed / heavy; grievous', whose dual valency is longstanding (attested in Don Juan Manuel, 14th Century (Alonso, 1494), following naturally from that of its verb, pesar being both transitive 'to weigh (something)' and intransitive 'to have weight'. The assumption of analogical pressure exerted within this field would give a natural explanation of these coincidences in valency.

Of these other past participles with double valency, pesado had long had the meaning of 'heavy' and from this meaning it no doubt extended figuratively to mean 'tedious, annoying'. It is attested in a number of such meanings in Autoridades, 241. It is in such meanings that it is chiefly used adjectivally, both attributively and predicatively, of both people and things. Cansado also appears to have gained an active valency relatively early: Autoridades, 119, cites an example from Alemán's Guzmán de Alfarache (16th Century). Covarrubias, 288, makes specific mention of the expression hombre cansado, defining it as 'el pesado en sus razones y trato, que cansa y muele a los que le han de tratar y conversar con él'. Cansado, unlike pesado, could be applied in both passive and active valencies ('tired' / 'tiresome, tiring') to animate beings. The rising availability in the Golden Age of the ser/estar contrast with adjectives[22] must have been a major factor in assuring the coexistence of both valencies, since it provided a means of disambiguation. This is very clearly demonstrated in Franciosini (1620) (quoted in Gili Gaya, 456):

(10)

It is likely, then, that cansado supplied the model for other past participles of the same semantic field. Autoridades, II, 522, records an active valency meaning for entretenido with an example from Quevedo (also in Requejo (1717) cited by Gili Gaya, 917). But still in Autoridades there is no attestation of either aburrido or divertido with active valency meaning; since Autoridades seems to be quite sensitive to this issue, regularly dubbing active past participles 'hispanismos', it is unlikely that such a valency would have been missed, and we must therefore conclude that the active valency of aburrido dates from relatively recent times. Apart from a 'rogue' medieval example, DHRAE's first recorded attestation in this sense is from Moratín (early 19th Century).
 
 

2.2.5 The semantic attenuation of aburrir

The second question is the semantic 'strength' of aburrir. The glosses given for aburrir in the 17th and 18th Centuries involve notions which are much 'stronger' than the modern meaning of 'bore'; even as late as the mid-19th Century we find Mora's Colección de sinónimos de la lengua castellana, cited in Cuervo, 80, charting the relative meanings of aburrir and fastidiar as follows:

(11)

It may be relevant here to consider the semantic history of another member of this constantly shifting semantic field: enfadar. According to Corominas & Pascual, II, 611-4, enfadar is a borrowing from Galaico-Portuguese. It makes a relatively late entry into Castilian, gaining currency in such writers as Cervantes, Góngora and Quevedo (it does not feature in Nebrija's Vocabulario). In Galician, it was always reflexive and had a meaning which was comparable with that of present-day Castilian aburrirse. This value is amply attested in Castilian too as late as the 18th Century; indeed, Corominas & Pascual point out that the 'stronger' meaning of 'enojar, encolerizar' is still uncommon in Argentina and Chile[23]. Thus Covarrubias, 519:

(12)

The attenuation in meaning of aburrir/aburrirse apears to go hand in hand with the strengthening in meaning of enfadar/enfadarse. Cuervo, 389-91, identifies two meanings for enfadar: (a) 'molestar, causar disgusto o tedio', and (b) 'irritar, enojar, enfurecer', which reflect the 'old' and the 'new' values respectively. The examples given for each seem to confirm that the popularity of meanings (a) and (b) are in inverse proportion, although there must have been a long period of coexistence of the two: the first attested example with meaning (b) being from Fray Luis de León (16th Century) and the last example of meaning (a) is as late as Galdós (19th Century). Enfadar and aburrir must at some point have come fairly close in meaning, as suggested by an 18th-Century example:

(13)

The interplay of enfadar and aburrir may therefore be of importance in the semantic attenuation of the latter, aburrir moving towards the old sense of enfadar as this in its turn strengthens towards the meaning of 'to make cross, annoyed'. It is probably only after the establishment of aburrir in this attenuated sense, and hence its semantic proximity to cansar, that the dual valency of aburrido is able to be developed, facilitated by the now firmly established ser/estar contrast.
 
 

3 Conclusion

The trajectory of ABHORRERE/ABHORRESCERE in Castilian may be summarised as follows. Medieval aborrescer/aborrir overlap semantically in many respects, but there is evidence to suggest that they were not altogether synonymous nor identical in syntactic possibilities. Nevertheless, it seems that so far as the finite verbs were concerned, there was an expected tendency during the Golden Age to prune aborrir from the language as an exponent of the concept 'hate'. Aborrir is eventually so pruned; but a form aburrir, (or at least a past participle form aburrido), which may have existed as a popular variant all along, survives in certain registers. The popular past participle evolves a specialised, attenuated meaning of 'dispirited, desperate' in addition to the meaning of 'hated', and a new valency is then transferred to aburrir, the latter being reinstated in the mainstream of the language with the meaning 'to cause to be dispirited, desperate'. This re-entry may well arrest the tendency to change valency that aborrecer temporarily shows. Aborrecer and aburrir are by the 18th Century more or less completely discriminated semantically, which is the key factor in the survival of what was originally an etymological doublet. The further semantic evolution of aburrir to the meaning of 'to annoy, bore' is achieved by the complementary movement of enfadar away from this semantic area. The graphic representation of the semantic evolution of aburrir given in (14) shows the evolutionary process as a series of related changes. Lastly, aburrido develops a double valency, facilitated by the strengthening of the ser/estar contrast, as a result of analogy with cansado and other close semantic relatives.

(14)

Footnotes

[1] Allerton, 2, for example, conceives valency very broadly as 'the capacity a verb has for combining with particular patterns of other sentence constituents'.

[2] This theme is explored further in Pountain (1993a).

[3] See Pountain (1993b).

[4] See Pountain (1993b).

[5] There are more complex relations still, which do not affect the material discussed in this article (eg hacer cantar 'to make to be sung', where the object of hacer can also be the object of cantar). Because in such a case the infinitive can be thought of as having a passive value, the valency of such an expression might be termed causative(passive).

[6] It does not seem uncommon for semantic evolution to involve valency changes: Gómez Torrego, 20, points to a number of changes in transitivity in modern Castilian, of which at least one involves a transitive non-causative -- causative(transitive) shift, the case of repercutir (en), which has extended its meaning from 'to have repercussions on' to 'to reflect in', ie 'to make to have repercussions on'.

[7] Cf. Ullmann, 233-4, who indeed illustrates 'ameliorative' development of meaning with the case of French ennuyer.

[8] Lewis & Short, 7.

[9] Latin ODI (ODISSE, OSUS) was very defective, existing only in the Perfect forms of the paradigm. There was every motivation for replacing this verb by a more regular form; but the supine does not in this case supply the model. See also Ernout, 172.

[10] The Vulgar Latin *INODIATUS which we take to be the origin of French annoié, Italian noiato, etc. cannot be identical to Classical INODIATUS (again presumed to be based on the noun ODIUM), which meant 'not hated'.

[11] Tobler-Lommatsch, III.1, 469-72.

[12] Devoto, 283.

[13] Cioranescu, 877.

[14] See Elcock, 127, 136-7, on the favouring of -ESCERE in Ibero-Romance and on the morphological developments of this form.

[15] Attested in the Vulgate, 2 Maccabees 6,12.

[16] I am indebted to Bosque & Pérez Fernández in establishing these pairs.

[17] Exceptions are agradecer 'to thank, be grateful (for)' / agradar 'to please' (and the corresponding negatives desagradecer/desagradar), embebecer 'to delight, fascinate' / embeber 'to absorb, soak up; shrink', tullecer/tullir 'to cripple' (both). Note how in all cases except the last the pairs have become semantically discriminated.

[18] Cf. Müller, 160: 'La redundancia léxica en este sector explica la pérdida de varias de estas formas en la diacronía de la lengua.'

[19] DHRAE, 249, treating aborrir and aburrir together, states that in the meaning 'despreciar, menospreciar...' aburrir is used intransitively, but the only likely example given involves a prepositional complement in de which according to my criteria is a transitive use.

[20] DHRAE's panchronic definitions 'sentir horror, aversión o disgusto' (248) and 'hacer sentir horror, aversión, disgusto o tedio' (250) make this relation very patent (but we must not ignore the fact that the semantic trajectory of aburrir does involve weakening).

[21] See DHRAE, 248.

[22] See Pountain (1982).

[23] Cf. Kany, 214: 'enfadarse "to be angry" is in America usually replaced by its synonym enojarse, since enfadarse seems to stress the marginal notion "to be bored" (aburrirse)'.
 
 

Bibliography

Allerton, D.J. (1982). Valency and the English Verb (London: Academic Press).

Alonso, Martín (1986). Diccionario medieval español (Salamanca: Universidad Pontificia).

Autoridades = Real Academia Española (1963). Diccionario de Autoridades. Edición facsímil (Madrid: Gredos).

Bosque, Ignacio & Pérez Fernández, Manuel (1987). Diccionario inverso de la lengua española (Madrid: Gredos).

Cioranescu, Alejandro (1966). Diccionario etimológico rumano (Tenerife/Madrid: University of La Laguna/Gredos).

Corominas, J. & Pascual, J.A. (1980). Diccionario crítico etimológico castellano e hispánico (Madrid: Gredos).

Covarrubias, Sebastián de [1611]. Tesoro de la lengua castellana o española (Madrid: Turner).

Cuervo, R.J. (1953-). Diccionario de construcción y régimen de la lengua castellana (Bogotá: Instituto Caro y Cuervo).

De la Porte (1659) = Den nieuwen Dictionaris oft Schadt der Duytse eñ Spaensche Talen... door Arnoldus de la Porte (Antwerp: Verdussen).

De las Casas (1576) = Vocabulario de las dos lenguas toscana y castellana, de Christoual de las Casas... (Venice: Damian Zenaro).

Devoto, Giacomo (1966). Avviamento alla etimologia italiana (Florence: Le Monnier).

DHRAE = Real Academia Española (1972-). Diccionario histórico de la lengua española (Madrid).

Elcock, W.D., revised Green, John N. (1975). The Romance Languages (London: Faber).

Ernout, Alfred (1954). Aspects du vocabulaire latin (Paris: Klincksieck).

Franciosini (1638) = Vocabulario italiano e spagnuolo...Vocabulario español, e italiano. En ésta Tercéra impresion corregido, y añadido por su verdadero Auctor Lorenzo Franciosini Fiorentino / Lorenzo Franciosin Florentin... (Roma: Camera Apostolica).

Gili Gaya, Samuel (1947). Tesoro lexicográfico (1492-1726). (Madrid: CSIC).

Gómez Torrego, L. (1991). 'La lengua española hoy' in España hoy: Cultura, ed. Antonio Ramos Gascón (Madrid: Cátedra), pp.9-37.

Kany, Charles E. (1960). American-Spanish Seantics (Berkeley and Los Angeles: University of California Press).

Lewis, Charlton T. & Short, Charles (1879). A Latin Dictionary (Oxford: University Press).

Minsheu (1623) = A Dictionary in Spanish and English: first published into the English tongue by Ric. Percivale Gent. Now enlarged and amplified...by John Minsheu... (London: John Haviland for William Aspley).

Müller, Bodo (1987-). Diccionario del español medieval (Heidelberg: Winter).

Nebrija = Gerald J. Macdonald (1973). Antonio de Nebrija. Vocabulario de Romance en latín. Transcripción crítica de la edición revisada por el autor (Sevilla, 1516) con una introducción. (Madrid: Castalia).

Oudin (1607) = César Oudin, Tesoro de las dos lenguas francesa y española... (Paris: Orry).

Oudin (1645) = Le tresor des deux langues espagnolle et françoise... (Paris: Sommaville, Courbé, De la Coste).

Oudin (1660) = Tesoro de las dos lenguas, española y francesa... (Brussels: Mommarte).

Pountain, Christopher J. (1982). '*ESSERE/STARE as a Romance phenomenon', in Vincent, Nigel & Harris, Martin (eds), Studies in the Romance Verb (London: Croom Helm), pp.139-60.

Pountain, Christopher J. (1993a). 'De la construçion de los verbos después de sí: la transitividad en la tradición gramatical española', in Ralph Penny (ed.), Actas del Primer Congreso Anglo-Hispano, Castalia, Madrid, pp.89-98.

Pountain, Christopher J. (1993b). 'Aspect and voice: questions about passivization in Spanish', in Journal of Hispanic Research, 1, 167-181.

Requejo (1717) = Valeriano Requejo, Thesaurus Hispano-Latinus (Madrid).

Sobrino (1705) = Dicionario nuevo de las lenguas española y francesa...por Francisco Sobrino...(Brussels: Foppens).

Tobler, Adolf & Lommatzsch, Erhard (1925-). Altfranzösisch Wörterbuch (Wiesbaden: Steiner).

Ullmann, Stephen (1970). Semantics. An introduction to the science of meaning (Oxford: Blackwell).