Abstract
Current research on the reflection of bodily states in language has opened the door to an investigation of the role that morphosyntax plays. In the case of emotions, these states have often been found in the lexicon and in metaphors, but morphology and syntax remain relatively unexplored. This study examines one emotion, fear, and the reflection of physical behavior in the morphosyntax of Indo-European languages. As fear invokes avoidance behavior, the language itself should indicate this desired spatial distance. Through discussions of case, mood/modality, and negation, this analysis illustrates avoidance behavior in morphosyntax through the linguistic conceptualization of distance.
1. Introduction
This article will investigate the marked morphosyntactic patterns of fear constructions in several Indo-European (IE) languages, specifically Latin, Greek, Sanskrit, English, and Baltic and Slavic languages. These morphosyntactic patterns will include the case marking that is required or selected with fear nouns and verbs, the choice of mood and/or modality, and the synchronic uses of the negative marker *nē̆.1 As this article will show, fear constructions in some IE languages are built upon the concept of distance. The natural behavior associated with fear is flight, which is the distancing of the Experiencer from the negative stimulus. This bodily behavior is reflected in the case markings, mood, modality, and negative particles that occur with fear lexemes.
This analysis will explore the morphosyntax of fear constructions in a number of IE languages. The goal is not a reconstruction of fear constructions in Proto-Indo-European (PIE), but rather a synchronic survey of fear constructions in relevant IE languages. It is also not meant to be comprehensive nor state that all languages, IE or otherwise, use these methods in their fear constructions. Rather it is meant to explore the patterns found in the languages which do use these markings, morphemes, and moods in their fear constructions. The use of morphological case marking will be discussed in section 2. In section 3, the relationship between fear and negation will be investigated with an emphasis on the genitive of negation in the Baltic and Slavic languages, as well as negative particles such as *nē̆/*mē in Greek, Latin, and French. Mood and modality in fear constructions are the topic of section 4, and the conclusion is presented in section 5. The analysis will show that the grammar of fear constructions in IE languages reflect the body’s response to fear through the linguistic conceptualization of distance.
2. Morphosyntactic Case Marking
Case marking can serve to relate the semantic roles within an utterance. As different case markings are related to different roles, basic level meanings for each marking can be proposed (cf. Jakobson 1936/1984 for Russian).2 This section will explore the various meanings that have been connected with the case markings that occur with common fear lexemes in IE languages.
Two case markings are associated with fear constructions in IE languages. These are the genitive and ablative, and an aspect of these case markings is the concept of spatial distance. The sections below will examine the use of these markings in PIE, as well as their synchronic uses in selected daughter languages. Although the choice of morphosyntactic markings may vary across languages, the semantic conceptualization of spatial distance remains remarkably constant.
2.1. Genitive Case Marking
IE languages that have retained a morphological case marking system often have genitive objects in fear constructions. Baltic and Slavic languages generally follow verbs of fearing with genitive objects, for example, as does Sanskrit in some instances. This use of the genitive in these languages can be explained when the basic meaning of this case marking is explored and interpreted in terms of a physical response to the emotion of fear.
The varied uses of the genitive usually prompt a listing of frequent functions in language grammars. In terms of its occurrence with fear verbs, it might be categorized as an aspect of the partitive use of the genitive, as the genitive of source, the genitive of origin, or the genitive of cause, depending on the individual language’s reference grammar. Some scholars, such as Jakobson (1936/1984) for Russian, have argued for a general meaning of morphological case marking that covers all the various functions each can have. Instead of detailing all the possible uses of the genitive in Russian, Jakobson condensed the different functions of the case into one meaning: limitation. The entity in the genitive is limited in its involvement in the events being discussed, and thus the genitive (G) is used to indicate “the limit of the referent’s involvement in the content of the utterance … The G in itself indicates only that the scope of its referent’s involvement in the content of the utterance is less than the referent’s entire extension … The referent of the G can be either partially or negatively represented in the sentence” (346–347).
Jakobson considered limitation to be central to the meaning of the genitive in Russian, and others (e.g., Watkins 1967) have extended this core meaning to all IE genitives. Watkins (1967: 2195) suggests that the possessive use of the genitive (or “genitive of belonging”) can be reconstructed for PIE, with examples in Old Hittite, Vedic Sanskrit, and Homeric Greek (2195). He further asserts that the subjective and objective uses of the genitive extend from this original possessive use (2198). How do we reconcile this original use of the genitive, possession, with Jakobson’s core meaning of limitation? Langacker’s (1993) discussion of possessives as reference-point constructions provides the connection. In a possessive construction, one entity (the possessor) acts as a reference point for establishing contact with the other entity (the possessed) (8).
Jakobson states that the genitive indicates the referent’s limited involvement, just as Langacker has the genitive case-marked entity as a reference-point, which entails its limitation in terms of attentional focus in the utterance. The definition of Janda & Clancy (2002: 110) for the genitive in Russian perhaps best exemplifies this limitation or reference-point status: “The genitive is a backgrounded item that yields focus of attention to something else which exists or maneuvers in its proximity.” Thus the concept of limitation becomes one of limited attention in the conceptualization of the utterance.
This core meaning of genitive case marking is present in all instantiations of genitive constructions in actual discourse per Jakobson’s (1958/1984: 107) concept of relational invariance which allows for an abstract meaning and the varying subset of meanings present in individual constructions. Following this line of thinking, the core meaning of genitive case marking is extended to domains beyond possession, domains which retain this invariant or prototypical meaning of limitation, but which have variant meanings as well, which are extensions of this invariant meaning. In terms of fear constructions in IE languages, one extension of the genitive case meaning in particular is present: the representation of SOURCE.
As noted above, genitive case marking has been related to source, separation, cause, and origin when grammars consider specific constructions. These can all be collapsed into the semantic domain of SOURCE. SOURCE is the starting point of the action or event, and, in spatial terms, genitive case marking denotes that the other entity is moving away from the object. Thus the concept of separation is inherent, as moving away from the source means a separation from it. Cause and origin can also be collapsed into SOURCE, as the cause and the origin in terms of an event is also the starting point from which the event moves away, temporally or spatially. For fear constructions, the use of genitive case marking indicates the same movement away, in this case movement from the fear stimulus. Limitation in terms of a behavioral response to fear would be the desire of the Experiencer to have distance between him/herself and the fear-provoking stimulus or SOURCE.
The Experiencer of a fear verb may not actually come in contact with the fear stimulus or SOURCE. Limitation is present in the fact that the fear stimulus may not exist or come to pass in reality (see section 3 below). Fear constructions can make use of these means (e.g., case marking, complementizer selection, and mood) to achieve the same function: marking distance. The sections below will examine the use of genitive case marking in IE fear constructions while exploring this concept of limitation in terms of physical reactions.
2.1.1. The PIE Genitive
According to current reconstructions, eight cases have been proposed for PIE: nominative, vocative, accusative, genitive, ablative, dative, locative, and instrumental (Beekes 1995: 173). Jakobson’s characterization of the general meaning of genitive case marking in Russian, that of limitation, has been extended to all IE languages, even to PIE itself. However, although the genitive case might be found in the daughter IE languages, the other oblique cases, namely the ablative, dative, locative, and instrumental, are not often all present in those languages. Instead, case syncretism has often taken place, in which one case ending absorbs the functions of another. Thus, while Latin has both the genitive and ablative, Greek has only the genitive, which also took on functions of the PIE ablative. For certain types of case syncretism to happen, the case markings which combine must be related semantically or functionally to motivate their combination (Baerman 2009: 219). In the case of the IE genitive and ablative, certain commonalities existed which prompted their combination in some languages. The examples below will illustrate the use of the objective genitive following fear lexemes in a selection of IE languages.
2.1.2. Genitive Case Marking
Many IE languages have retained their case marking systems, and many still use morphological affixes to mark grammatical relations within a clause. Several Baltic and Slavic languages, for instance, have retained a productive system and these languages often use the genitive when marking the object of a fear lexeme (see section 3).
As can happen over time, case syncretism has occurred in some IE languages. For example, the ablative merged with the genitive in both Old Church Slavonic (OCS) and Ancient Greek, leaving only the genitive. However, particular uses of the genitive are often referred to as “ablatival genitives” (Smyth, 1920: 328 for Greek; see also Schwyzer 1950: 90–101, who provides a comprehensive list of the ablative functions of the Greek genitive, inter alios). Under the ablatival genitive fall some verbs of emotion. Grammars such as Smyth’s (1920: 328–332) for Greek group many of the verbs that occur with an ablatival genitive complement under headings such as “Genitive of Separation,” “Genitive of Cause,” and “Genitive of Source.” Genitives of this kind include those accompanying verbs that signify ceasing, releasing, removing, restraining, giving up, failing, being distant from, depriving, lacking, wanting, and emptying (328–330). They also include emotional verbs such as “to wonder at, admire, envy, praise, blame, hate, pity, grieve for, be angry at, take vengeance on, and the like” (330). These categorizations indicate not only that the genitive and the ablative cases have similar semantics, but also that their semantics reflect bodily experience in perception and spatial distance. The genitives of separation, source, origin, and cause all involve the perception of distance, physical or conceptual, between the Experiencer and the complement. The “separation” verbs involve acknowledging the distance between the Experiencer and the complement, perhaps in stopping an action and therefore moving “away” from the SOURCE in a conceptual sense, or in no longer or perhaps never having the complement, in the sense of deprivation verbs.
The core meaning of the genitive is very similar to that of the ablative in terms of conceptualization; therefore, both the genitive and the ablative can be construed to have similar conceptualizations in terms of the response to fear, namely movement away from the fear stimulus. The following section illustrates the use of the ablative in IE languages, and its correspondence to the genitive in core meaning and function.
2.2. Ablative Case Marking
The ablative shares many similarities with the genitive, and, as noted above, shared semantics between the two likely led to syncretism in some IE languages. However, other IE languages such as Sanskrit and Latin retained both the ablative and the genitive as separate case markings. While the genitive was used to indicate possession and its polysemous meanings, the ablative indicated movement away from SOURCE. Like the genitive, the ablative has an associated list of functions such as the ablative of origin, privation, and want, the ablative of source, the ablative of cause, and the ablative of separation (examples here given from Latin; see Allen & Greenough 1903). Once again, with movement away from SOURCE, limitation is present in not only the backgrounding of the ablative-marked referent in terms of the other argument of the utterance, but in the limiting of the referent’s interaction. Movement away from SOURCE in terms of emotional verbs such as ‘fear’ mimics the bodily reaction of flight and the desire to put distance between the Experiencer and the source of the fear emotion.
2.2.1. The PIE Ablative
PIE had both the ablative and genitive, with the ablative generally indicating origin, and the genitive possession. Nevertheless, these case markings might have still had overlapping semantics due to the underlying conceptualization of each. The genitive is tied to limitation, but this characterization can also be extended to the ablative as both limit the interaction of the referent, often spatially. In fact, it is commonly believed (see, for example, Beekes 1995: 91) that a partial case syncretism took place in PIE for the genitive and ablative cases. Support for this view comes from Sanskrit, where the same overlap occurs.
2.2.2. Sanskrit Ablative Marking
The ablative in Sanskrit generally indicates the source or starting-point (MacDonell 1927: 190), and with verbs of fearing, the ablative is used to indicate the SOURCE. Again, as with the genitive, the ablative indicates distance between the Experiencer and the fear stimulus.
(1)“everything was afraid of her at her birth” (Whitney 1889: 97)
However, genitive case marking could also be used a verb of fearing, e.g., bhī-:
(2)“we are afraid of thee.” (Whitney 1889: 100)
The semantic relationship between these two case markings seems to hinge on spatial conceptualization, of which fear constructions reflect one instantiation. Both indicate the limitation of the SOURCE in its involvement with the fear Experiencer. The concept of limitation is inherent in this conceptual distance, as the speaker attempts to limit his or her interaction with the SOURCE, and to separate him/herself from the fear stimulus (physically and conceptually).
2.3. Genitive and Ablative in Fear Conceptualization
While this analysis puts forth the theory that the concept of spatial distance is mirrored in the morphosyntactic constructions used with fear lexemes, this notion can be further refined and connected to methods of conceptualization. As fear prompts the Experiencer to limit his/her interaction with the fear stimulus, as well as to possibly seek a spatial separation from the stimulus, this separation can only be accomplished by movement away from SOURCE. The fear stimulus could be marked by the genitive or ablative case because both are associated with SOURCE in many IE languages.
Another aspect that is often present within fear constructions is negation. The section below discusses negation in terms of fear constructions, from negative semantics to specific negative particles and their use in complex clause structure. It will illustrate that negation provides another example of the reflection of the physical reaction to fear in language.
3. Negation
Strong negative emotions such as fear are distinct in the sense of both their affect and their conceptualization. Fear evokes a strong response, perhaps one of the strongest among emotions, and the response to fear can be considered negative in the sense that the Experiencer has negative feelings (as opposed to emotions considered generally positive, such as happiness and joy) and an uncontrolled change in one’s bodily state (Rozin & Royzman 2001). Fear constructions in IE languages can mirror this negativity in their morphosyntax through their selection of case marking and the use of negative complementizers.
3.1. Baltic/Slavic Languages and the Genitive of Negation
The use of the genitive with verbs of fearing and avoidance in Baltic and Slavic languages is an old construction as it is attested in both Old Church Slavonic (OCS) and in both modern Baltic languages. It should be stressed again that the focus here is on a specific function of genitive case marking in these languages indicating avoidance. Other verbs with both negative and positive semantics can make use of the same case marking for a different function, just as the same functions of the genitive, such as indicating possession, can be expressed with other case markers, such as the dative. The genitive of negation, in which the genitive is used for the complement of a negated transitive verb, existed in OCS, just as it continues to exist in Baltic and most Slavic languages today. As the examples below will show, the semantics of fear and avoidance verbs lend themselves to a negative categorization, despite the fact that the verbs themselves are not actually negated.
The verb category under which the subset of fear verbs falls, furthermore, generally denotes not only limitation, but also, more specifically, the conceptualization of distance on the part of the speaker. However, there are two senses of distance with emotion verbs. As will be seen in other IE languages that make use of genitive case marking (namely Baltic and Slavic languages), a similar subset of verbs occur including those meaning ‘seek,’ ‘search for,’ ‘await,’ ‘demand,’ ‘want,’ ‘reach,’ ‘achieve,’ and ‘acquire,’ but also those meaning ‘be ashamed of,’ ‘mind,’ and ‘avoid.’
Timberlake (2004) makes a distinction among the semantics of each type of these verbs for Russian, stating that the genitive object is one that is only potentially affected, and the contact between subject and object is only potential and not actual. Thus, although the same case marking can be used for both subsets of verbs, and though limitation, the core meaning of the genitive, still applies, the distance between the two types of verbs and their complements is the result of a difference in conceptualization. For each of these verbs that take genitive case marking, distance is present. This distance is connected to the concept that the interaction has not yet taken place and may not at all. The interaction between the subject and the object is possible, but not yet actual. For avoidance verbs, including fear, the distance between the Experiencer and SOURCE is one that is purposely desired, while for the other subset of verbs, the distance indicates the potential of contact that has not yet occurred, but could possibly be welcomed, unlike the distance accompanying avoidance verbs.
3.1.1. Old Church Slavonic
In Lunt’s grammar of OCS (2001: 145), the verb bojati sę ‘be afraid of’ takes a genitive object, as do verbs such as lišiti ‘deprive,’ stradati ‘suffer loss of,’ sramljajǫtъ sę and postyděti sę ‘be ashamed of’ and běžati and běgajǫtъ ‘flee from.’ These verbs take their complements in the genitive (into which the ablative of PIE had collapsed), and although these verbs are not negated in actuality, their semantics are decidedly negative.
Not only are these semantically negative words, but fear verbs also include the notion of the desire to prevent the actual event. As Jakobson (1936/1984: 347) indicates for Russian, the genitive referent may be partially or negatively represented, and much of the literature discussing the genitive in several Baltic and Slavic languages revolves around this negative representation, often referred to as the genitive of negation. Scholars have developed many semantic and syntactic theories to account for this behavior, attributing the use of the genitive of negation to the individuation of the participant (Timberlake 1975), the scope of negation (Babby 2006), perspective structure, meaning whether the event is framed from the perspective of the object or the location (Partee & Borschev 2004), unreality (Neidle 1988), or “the absence of commitment to existence” (Kagan 2010: 21).
These interpretations of the function of the genitive of negation can be enfolded within Jakobson’s general meaning of limitation. Negation with a lack of individuation necessarily entails a limitation on the object’s involvement in a negative sentence. Furthermore, discussions that hinge on existence also can be brought within limitation, as the absence of existence (whether of the thing itself or its existence at a location) means that it must have a limited involvement. Jakobson’s general meaning of the genitive as limiting a referent’s involvement in the event is evident with fear verbs, as the limitation is also imposed by the Experiencers themselves in their desire to avoid the possible negative event. Thus, in OCS, as in many Baltic and other Slavic languages, fear verbs and their counterparts seek to move away from the possible negative stimulus that is contained in the genitive complement, instantiating the conceptual distance between the Experiencer and the fear source which mirrors the physical distance of the fear response.
3.1.2. Russian
In Russian, as in the other Slavic languages that retain their synthetic morphology, verbs of fearing or avoiding generally take genitive case marking for their objects (as a subset of intentional verbs).3 Verbs in Russian that require the genitive include bojat’sja ‘fear/be afraid,’ izbegat’ ‘avoid,’ and osteregat’sja ‘beware.’ As in OCS, this use of the genitive can be considered a part of the overall semantic domain of the genitive of negation as the negative semantics of these verbs prompt the genitive. The genitive of negation is itself a reflection of the meaning of limitation, which in the case of fear verbs also reflects the desire of the speaker to move away from the fear stimulus, physically or conceptually.
(3)“I’m afraid of the dark.”
3.1.3. Lithuanian and Latvian
Lithuanian is considered to be, in many respects, the most conservative of the modern IE languages. Not only does the genitive of negation exist in Lithuanian, but the genitive is also used for verbs of striving and avoidance, just as in the related Slavic language family. As in Slavic, verbs with meanings such as ‘long for,’ ‘wait for,’ ‘wish for,’ ‘want,’ ‘hope for,’ and ‘yearn for’ all take a genitive complement (Mathiassen 1996: 184). While these verbs have the notion of positive expectation, verbs with the opposite notion also take the genitive. These verbs include bijóti or baidýtis ‘be afraid,’ išsigãsti ‘be frightened,’ gė́dytis ‘be ashamed about,’ sáugotis ‘mind,’ and šãlintis or véngti ‘avoid.’ The semantics of all of these verbs contain the notion of not yet achieving or confronting the object of the clause. While some have a positive sense of expectation (e.g., ‘hope for’), others, such as ‘be afraid’ or ‘avoid’, anticipate an object the Experiencer views negatively.
Latvian, the other modern Baltic language, behaves similarly. Latvian verbs that can take a genitive complement could be considered under the categorization of deprivation once again. These verbs include alkt ‘crave,’ kārot ‘desire,’ gaidīt ‘wait for,’ meklēt ‘search for,’ vajadzēt ‘need,’ vairīties ‘avoid, evade,’ and, of course, baidīties ‘fear’ (Nau 1998: 24). The semantics of these verbs again hint at the lack or possible nonexistence of something, as well as the desired separation from something, and thus fall under the wide semantic category of limitation.
What these verbs have in common in their case marking is the notion of distance, marking the Experiencer as spatially distant from the SOURCE of the emotion. This spatial distance may not be actual physical distance, as a separation of an emotion from the person who is experiencing it is perhaps impossible, but the genitive case marking’s function of indicating spatial distance can be extended to indicate conceptual distance, or a desired separation from the source of fear. As in Slavic languages, this subset of verbs share the same marking, but with different shades of meaning. Whether the Experiencer desires to move towards or away from the stimulus of his desire, there is an implied distance between the SOURCE and the Experiencer. For verbs of fearing, the Experiencer seeks to limit interaction with the fear stimulus, so that perhaps no part of the object feared is experienced.
The genitive of negation, by its reference to an interaction that is limited in that the genitive object is either partially or negatively represented, can be connected to the irrealis mood, which will be further discussed in section 4. As Kagan (2010: 2013) states, there is a commonality between subjunctive mood clauses and certain genitive objects (which she terms “Irrealis Genitives”) in their absence of commitment. For the irrealis mood, the absence of commitment is to the truth of the proposition in the subjunctive mood clause, while the genitive can signal the absence of commitment to existence in that the genitive-marked object might not exist in the real world. In the case of fear constructions, the Experiencer may not actually come in contact with the fear stimulus. By the same token, irrealis mood is also used when the proposition may not come to pass, but is only possible or probable.4 Both markings can thus be connected by their similar functions.
Morphological case marking is not the only method of marking the negative semantics and distance within a fear construction. While genitive case marking in the languages mentioned above is one method of illustrating limitation with fear lexemes, other IE languages use marked particles to demonstrate this distance. One particle, in particular, negator *nē̆, is used in several IE languages with fear and related constructions.
3.2. Indo-European Clause Structure and *nē̆/*mē
Another commonality across several IE languages is the development of subordinators or complementizers to introduce subordinate clauses. Several of these subordinators developed from particles in PIE, especially the negators *nē̆ and *mē.
Through the comparison of many daughter languages, two negative particles have been reconstructed for PIE: *nē̆ and *mē. According to Lehmann (1974: 124), *mē was used in prohibitions, while *nē̆ was used elsewhere. As a comparison of several IE languages illustrates, the prohibitive made use of a separate particle (sometimes accompanied by an unmarked ‘injunctive’ form of the verb, or the verb in another mood, imperative, subjunctive, or optative). Several IE languages, including Anatolian, Indo-Iranian, Greek, Latin, Tocharian A, and Armenian each used a different negative particle for a negative prohibition and a negative statement (see Table 6.1 in Clackson 2007: 163).
Prohibitions in PIE were made with a special verb form, often called the injunctive, which may have originally been unmarked for tense or mood (Meier-Brügger 2003: 255–256; Clackson 2007: 132). The injunctive could be used to express both prevention (to stop an action before it happens) and inhibition (to stop an event in progress).5 After the development of the imperative, in certain daughter languages the use of the negative particle with prohibitions could have been reanalyzed as a “modal negative” which was used with certain moods to make negative commands (Clackson 2007: 164).
This particle was carried down into the daughter languages, either as the sole negative particle (as in the early period of Latin), or as part of a pair, similar to PIE (as in Greek). It began to make its appearance in fear constructions attested in these daughter languages. As the following sections will illustrate, the use of a subordinator and complex clause structure in fear constructions is another means of expressing distance.
3.2.1. Latin nē
In Latin, one of the PIE negative particles became the sole negative particle during the language’s early development. Latin nē became relegated to use with non-indicative moods, and later introduced subordinate clauses, as another negative particle, nōn, arose to take over the duty of direct negation. Hence nōn began to take over the job of PIE *nē̆, leaving the latter to appear mainly in modal constructions.
Not only can nē be used in independent clauses of non-indicative mood, but it can also act as a subordinator in certain dependent clauses. These clauses all belong to the same semantic domain in which the function of nē and the subjunctive mood is to indicate the speaker’s desire to limit his or her involvement with the event in the subordinate clause. Like the use of genitive/ablative case marking, the use of the negative particle was another means of indicating a desired distance between the Experiencer and the SOURCE or stimulus of the fear emotion, which now takes the shape of a subordinate clause and the event it contains. The clause types that make use of this negative particle include negative purpose clauses and hindrance clauses as well as fear constructions.
(4)Negative purpose clause:
“… so that he might not decide against the Republic” (Sallust, Cat. 26)
(5)Hindrance clause:
“For our narrow places prevent the multitude of the enemy from being able to surround us.” (Caesar, B. Afr. 13)
When fear constructions in Latin are compared against these constructions which contain the negative particle nē, the pattern that would be expected is I fear that X may not happen. Instead, we find the pattern I fear that X may happen. In fact, to negate a fear construction, another particle, ut, must be used, or the negator nōn must be added to nē. The examples below illustrate the difference between non-negated and negated fear constructions.6
(6)Non-negated:
“I am afraid that she might see me.” (Morwood 1999: 102)
(7)Negated:
“I am afraid that she might not see me.”
Why use a negative particle to express an idea that is not grammatically negated? The answer may be that the semantics contain some inherent notion of negativity. In the realm of fear constructions, emotion is particularly salient as the strong negative emotion of fear in the main clause is being triggered by the event in the subordinate clause (or source of fear). Fear not only signals an emotional response, but also a conceptual distance between the Experiencer and the source in the desire to avoid or prevent the potential situation in the subordinate clause. Thus, for this occurrence of nē, although it does not syntactically negate the following clause, it does signal the underlying semantic domain.7 This same negative domain of avoidance/prevention is shared by nearly all nē constructions, both independent and dependent.
3.2.2. Greek mē
Ancient Greek had two negators, ou and mē. Like nē in Latin, mē in Greek could be considered a modal particle, as it is used in subordinate clauses in situations involving emotions, commands, prohibitions, wishes, oaths, purpose, apprehension, and fear (Smyth 1920). Complex fear constructions in Greek take mē as the complementizer introducing the subordinate clause.
(8) kaì hoi Athēnaȋoi, … phoboúmenoi mè̄ sphísi dícha gignoménois rhȃion máchōntai
“and the Athenians, fearing lest they get divided and so fighting at a disadvantage” (Thucydides 6.100, cited after Kitis 2009: 425)8
As in Latin, mē does not actually negate the verb in fear constructions, and if the subordinate clause is to be negated, it requires the other negator following mē. Modern Greek retains two negators, as well, with mi(n) and δen; the former follows the subjunctive marker na, and is used in similar constructions. Thus, Greek also uses the subjunctive mood, subordinate clause structure, and a specialized complementizer to indicate the distance between the speaker’s desire and the event in the dependent clause.
3.2.3. French ne
Although both Latin and French use a similar negative particle to introduce their subordinate clauses, these particles are not identical. In French, the ne that occurs in fear constructions is referred to as the expletive negator. In contemporary French, negation is generally accomplished with the ne … pas formula. However, in colloquial French, the ne is often deleted or reduced. And yet, in fear constructions in formal French, ne occurs alone, and instead of signaling the negation of the subordinate clause, it acts as a complementizer for the subordinate clause but does not negate it (akin to Latin nē).
(9)“I am afraid that he may come.” (Zeijlstra 2004: 65)9
This type of construction requires the subjunctive mood for the subordinate clause in French, as in the Latin examples above.10
The use of a particle with unique semantics as well as of an irrealis mood serves to indicate that this construction is distinct in multiple ways, which makes it further away conceptually from more typical constructions. This conceptual distance mirrors the physical distance of the fear response. The use of expletive ne is triggered by the negative semantics of the utterance and can be found in clauses expressing hindrance, doubt, and prohibition, as well as fear. The function of this expletive ne is to express conceptual distance between the clauses, which is shown by the use of a negative particle that differs from the one used in simple negation.
Negation in fear constructions is driven by a unique semantic conceptualization, as the subsections above have attempted to illustrate. Case marking and constructional semantics in fear constructions are tied to the use of a particular negative particle, although the clause introduced by the negative particle is not being negated. Verbs with similar semantics also require the use of the negative particle when nothing is being negated. Although the nuances of each language are different, verbs meaning ‘deny,’ ‘avoid,’ ‘fear,’ ‘refuse,’ ‘prevent,’ and ‘doubt’ can trigger the occurrence of the negative particle as well. All of these verbs share negative semantics, and it is these semantics that motivate the use of the particle, not actual syntactic negation.
The preceding sections have argued that avoidance constructions, including fear, are distinct in their morphosyntax, just as the concept of negation itself is distinct. Negation is, therefore, one of the methods of indicating a shift in conceptualization, as is the use of a certain case marking. Another method of illustrating this conceptual distance is through the use of an irrealis mood.
4. Mood and Modality
Fear predicates followed by object complement clauses often take an irrealis mood in some IE languages, generally the subjunctive. In other languages, such as English, in which the subjunctive mood has been declining in use, these complement clauses often contain modal auxiliaries which can serve a similar function. Furthermore, the range of functions of fear constructions has changed. Initially fear constructions indicated that a negative event might take place. However, the domain of fear constructions has expanded to include a more subjective function, one which expresses the speaker’s desire to distance himself from the negativity associated with the proposition present in the subordinate clause (see Jing-Schmidt & Kapatsinski 2012, for a discussion of this function of fear constructions). The following sections discuss mood and modality in relation to fear constructions and their encoding of a physical response in IE languages. Mood will be used to describe the morphosyntactically-marked category, while modality will refer to the speaker’s attitude toward the world, and includes notions of probability, necessity, and reality.11
4.1. Mood in PIE
PIE is thought to have had five moods: the indicative, imperative, injunctive, optative, and subjunctive. Four of these moods, the imperative, injunctive, optative, and subjunctive, were marked by grammatical inflection. This leaves the indicative mood unmarked grammatically. As argued by Givón (1979: 55), the declarative sentence is the default sentence type, due not only to it being the most frequent sentence type in discourse, but also due to its lesser morphosyntactic complexity, fewer distributional restrictions, and earlier acquisition. The function of the indicative is to state facts or express reality as it exists, without providing the speaker’s interpretation of or feeling towards these facts. The other moods, in contrast, function to express the speaker’s attitude or desire towards the proposition he is making. For the imperative the speaker is expressing his will to have some action performed, or his obligation to perform some action. In the case of the optative and the subjunctive, the mood often involves the speaker’s attitude towards the proposition.
The optative mood is generally described as the mood of wishes and desires. The subjunctive mood, on the other hand, is the mood of possibility. Both, thus, provide the speaker’s attitude towards the proposition being made, whether indicating the speaker’s desire that something take place or the desire to avoid an event the speaker feels may possibly take place. These moods are marked not only grammatically, but also cognitively in that they indicate that a statement is not expressing reality as it is, but as it could or should be. The function of these marked moods, then, could be to highlight their difference from the indicative, or the real. These moods, in particular the optative and subjunctive, indicate a conceptual distance between what is and what might be. The use of one of these marked moods in fear constructions to indicate conceptual distance is an element of the overall concept of distance.
Both the optative and subjunctive moods are reconstructed for PIE. The functions of the optative were to express a wish (the desiderative use) or possibility (the potential use), while the subjunctive functions were either prospective or voluntative (Meier-Brügger 2003: 257–258). The prospective subjunctive reported expected future happenings, while the voluntative subjunctive indicated the will of the speaker. In terms of the speaker’s will, the difference between the optative and subjunctive seems to hinge upon whether the speaker has the ability to bring about the desired action. When making a wish in the optative mood, the speaker cannot directly bring about his or her desire, while a first-person subjunctive usage would be an action within the speaker’s power.12 Although distinct functions exist between the moods, it is clear that there is some functional similarity in that both can express desire as well as possible future events.
In most of the IE daughter languages, the optative and subjunctive did not survive as separate moods.13 Generally, only one mood was kept with the functions of both the optative and subjunctive being collapsed in a case of what could be called mood syncretism. Thus, for example, Classical Sanskrit keeps only the optative mood, while Modern Greek retains only the subjunctive. Because the functions of both moods are related, their collapse into a single mood follows the maxim that similarity of function leads to similarity of form.
A related aspect of these moods is their use with future events. Because an event has not yet taken place, it still technically exists in the realm of the probable or possible as it is not yet actualized. Events that are desired, hoped for, awaited, and even feared, have generally not yet come to pass. Could the distance represented by the subjunctive or optative mood, then, not just be a temporal distance? After all, these moods are generally productive and are used in many constructions besides ones dealing with fear. Thus, the question becomes, how do we know that the function of these moods/modals in fear and related constructions is not just to indicate a possible future event?
The use of the irrealis with fear and avoidance constructions might reflect an earlier, semantically conditioned use of these moods. Whether the use of the subjunctive mood in fear constructions evolved originally from its main-clause agent-oriented use into speaker-oriented modality expressions such as commands, obligations, wants, or desires, or from a semantic concord between the epistemic modality in the main and subordinate clauses, either would still be present before the use of these moods/modalities could be extended to less semantically congruent and more syntactically conditioned contexts, such as solely indicating future events.14 Therefore, the conceptual distance present in these clauses depends on intention as it relates to positive or negative desire (which in this context is fear). If the emotion is positive or reflecting desire, the Experiencer could be seen as wanting to move through that distance towards a goal. However, if the emotion is negative or reflecting fear, the Experiencer could be seen as wanting to avoid the negative event and it is the preservation of this distance that is desired. Regardless of the Experiencer’s desires, the distance between the SOURCE and the Experiencer remains conceptually.
4.2. Subjunctive Mood and Epistemic Modality
The subjunctive mood can have many functions, often called “idiomatic functions” (see Allen & Greenough 1903’s list for Latin in §438, for example). While many grammars may list these functions without appealing to any unitary definition of the mood, in general it can be said that the subjunctive expresses some form of unreality when contrasted with the indicative. The subjunctive mood is the mood of hypotheticals, of possibilities and probabilities, as well as the means of expressing epistemic modality. Because it has a range of seemingly unrelated functions, often the subjunctive is considered to be conditioned syntactically and semantically empty. However, in fear constructions, the semantics that condition the use of the subjunctive remain evident and are not empty, but rather are reflective of the conceptual distance between the SOURCE of fear (now in the subordinate clause) and the Experiencer (in the main clause), a distance the Experiencer desires.
Epistemic modality, also known as propositional modality (Palmer 1986), is generally defined as the indication of the speaker’s commitment to the truth of the proposition (Bybee et al. 1994: 179–180). This definition can be further refined to include the speaker’s attitude toward the proposition, regardless of its truth or possibility. In terms of fear constructions, commitment to truth or the expression of uncertainty does not seem to fully describe their function. Instead, the speaker is often fairly certain that the unwanted event may take place, and therefore the truth or probability could be in little doubt. However, because the speaker’s attitude toward the proposition (which is negative in fear constructions) is highlighted, fear constructions can still be included under epistemic modality.15 This definition of epistemic modality has expanded over the years to indicate fear on the part of the speaker of giving possible offense to the listener, although such pragmatic uses are outside of the scope of this paper (but see Jing-Schmidt & Kapatsinski 2012 for fuller discussion).16
In typical complex fear constructions the speaker expresses anxiety toward the proposition in the subordinate clause, and the function of the subjunctive could be to indicate distance, like the other markers employed in fear constructions. An irrealis mood is used with events that are probable but not yet actual, and in that sense it is marked, or more conceptually distant, than expressions using the realis (indicative) mood. In fear constructions specifically, this conceptual distance is reinforced by the speaker’s own desire to impose a distance between him/herself and the fear-inducing event. The behavioral response of actual spatial distance is represented in the conceptual distance the subjunctive mood expresses. Grammatical distance, therefore, stands in for conceptual distance, which is an extension of the concept of spatial distance in the real world (see Haiman 1983 and Kirsner 1985).
In the languages discussed in the next sections, the subjunctive is generally used with fear constructions.17 The examples below highlight the use of the subjunctive to indicate conceptual distance between reality and the speaker’s will.
4.2.1. The Latin Subjunctive
In Latin the subjunctive mood took on the functions of the optative, as well as being used for polite imperatives. While the subjunctive could be used in main clauses to indicate commands or wishes, many of the constructions that require the subjunctive are subordinate in nature. Complex fear constructions in Latin require the subjunctive mood in the subordinate clause, which contains the anxiety-provoking possible event. The descendent Romance languages have also preserved the subjunctive mood for similar functions, although its use with fear predicates is not always required (e.g., fear predicates can occur with the indicative or the subjunctive in Spanish). These Latin fear constructions are not alone in their use of the subjunctive, however, or their use of the complementizer nē. In fact, those Latin constructions that take nē and the subjunctive mood in the subordinate clause can all be considered members of the same semantic domain, namely avoidance.
It may be expected that fear constructions belong to the domain of avoidance, as this mirrors the actual physical response to fear stimuli. The examples below illustrate this desire for avoidance. In (10), the Experiencer (grammatical Subject) in the main clause fears that he might be considered a co-conspirator in the wrongdoing mentioned in the subordinate clause:
(10)“And he underwent the fear that the convicted one might drag him in as the author of the nefarious deed.” (Livy 45.5)
The fear experience in (10) is provoked by the possible negative future event in the subordinate clause and the Subject’s desire to avoid that event. However, this type of construction is not limited to expressions of actual physical fear, but can be used to indicate social anxiety, as well.
(11)“He began to fear that I might be angry with him, and he was not mistaken: I was angry.” (Pliny, Ep. 1.5)
The context of (11) above is a letter written by Pliny the Younger complaining about the deeds of a political and social rival. It is unlikely that this rival actually felt any physical fear of Pliny’s annoyance. It is more likely that this rival is anxious to protect his social standing. This use of the fear construction seems to straddle the line between actual fear and an affective stance marker. Nevertheless, whether the desire to avoid an event provokes an actual physical fear response or just general anxiety, this construction functions to indicate the desired distance. By using the grammatical distance implied by the subjunctive, which is marked in terms of mood, this construction is able to replicate the fear response of avoidance behavior through Latin morphosyntax.
4.2.2. Baltic/Slavic Moods
Baltic and Slavic languages are inconsistent in their use of mood with subordinate clauses. The Baltic languages Latvian and Lithuanian can use the subjunctive mood in their dependent clauses, such as those following fear predicates, but such a use may not be required.18
(12)Lithuanian:
“He was afraid that she would get cold.” (Mathiassen 1996: 133)
Slavic languages, on the other hand, do not make use of the subjunctive mood with predicates of fearing. Although they have both a conditional form and a subjunctive form, the indicative is used in fear constructions, as with other verbs of perception, cognition, and speaking. As is illustrated in (13) by Russian, the indicative mood occurs with the verb meaning ‘fear’ in the Slavic languages.
(13)“He was afraid that she would get cold.” (example courtesy of C. Vakareliyska)
This section is by no means comprehensive, but is intended to illustrate that these non-indicative moods are used in some IE languages to indicate possible events to distance them from actual ones, and that the subjunctive form is generally used in fear constructions, where it functions to indicate not only negative futurity, but also conceptual distance.
4.2.3. The English Subjunctive
The use of the subjunctive mood in English has been declining and is rarely used in contemporary conversational American English. In Old English, verbs could be inflected for the subjunctive mood. In terms of complex fear constructions, the main verb of the subordinate clause often took the subjunctive when it contained a potentially negative event. The complementizer lest often occurred with these subjunctive subordinate clauses.19 Lest arose from the Old English expression ðy læs ðe ‘whereby less that.’ The phrase itself is comprised of the instrumental of the demonstrative pronoun ðy, followed by the adjective læs ‘less,’ and ending with the relative particle ðe. This phrase is first attested around 1000 and is used to introduce a clause containing a possible negative future event that should be watched out for or prevented (OED on-line).
Lest, though it does not negate anything, is semantically negative like the complementizers discussed above. Languages that have lexemes such as lest in English generally make use of them to signal an undesirable situation that should be avoided. Lichtenberk (1995: 298) points out that these elements can have avertive or precautionary functions. In fear constructions, as well as negative purpose clauses, lest frequently occurs with a verb in the subjunctive mood. Together the morphology of the verb and the syntax of the complex clause structure signal the negative semantic content and desired avoidance inherent in fear behavior.20
English fear constructions contain the subjunctive form or subordinate modal verbs in order to not only invoke a future possibility, but also to show obligation and the speaker’s stance. Whether the fear construction signals an actual physical fear or social anxiety, the pairing of modal verbs and subordinate clause structure combine to indicate the speaker’s desired distance between her/himself and the event in the subordinate clause, or the social offense that might be caused by the contents of the proposition.
If this analysis is correct, it follows that semantically distinct elements are used in order to reflect the body’s fear response through the conceptualization of spatial distance, depending on whether the fear construction occurs in a simple or a complex clause. In simple clauses, the morphosyntax of the genitive or ablative provides the notion of limitation that mirrors the Experiencer’s desire to limit his interaction with the SOURCE. In more complex clause structure, negative particles and certain moods and modalities echo the bodily limitation and separation through conceptualization.
5. Conclusion
As this analysis has attempted to show, elements of human physical experiences can be reflected within the languages we speak. In terms of fear, one of the strongest and most basic of human emotions, the reaction to negative stimuli leads to an avoidance response, which generally results in the experiencer distancing himself from the stimuli. This avoidance behavior can be reflected in morphosyntax. A language may use any one or more of these strategies. As was shown above, Lithuanian uses all three. This analysis focuses solely on IE languages, and the examples discussed above are by no means exhaustive, as other IE languages exhibit similar characteristics. However, it is likely that this reflection of the physical response to fear can be found in the morphosyntax of non-IE languages as well, and this is an area in need of further research.
The conceptualization of spatial distance that is found within IE fear constructions is ideally suited to model avoidance behavior. There are several ways in which this distance is expressed in the morphosyntax of fear constructions. It can be found in the use of certain case markings associated with spatial distance (i.e., the genitive and ablative). It is also present in the use of certain moods, modalities, and complementizers. Because the subjunctive mood is an irrealis mood indicating non-reality, it can function as a distancing element. Both the case marking and the mood choice revolve around this notion of absence of commitment to the existence of the object or event in reality. In this sense, both can be considered irrealis in that the object being feared may not be encountered in the real world, just as the event in the irrealis mood may not come to pass. Negation itself can be considered irrealis along the same terms, as the negative event may not exist in reality. The choice of a particular complementizer to introduce clauses dependent on fear predicates can also reflect this concept of irrealis, then, as the selected complementizers can be associated with the desire to avoid a possible negative event so that it might not come to pass. In each case, this sense of distance can be reflected in grammar through these irrealis-related methods.
Recent research in linguistics has been focusing more closely on the reflection of bodily states in language. Although much work has been done on lexical items and metaphors, less focus has been given to the actual morphology and syntax of specific languages. Work on emotional language is similarly limited. Further research is needed to explore whether other emotions show similar aspects of bodily states. Nevertheless, the preliminary research presented here is a first step towards connecting the physical world to our conceptualization and expressions. Avoidance behavior leads to spatial distance, and, in several IE languages, this distance is evident in the morphology and syntax of fear constructions.
References
Babby, Leonard. 2006. The Genitive of Negation and Unaccusativity. In Negation in Slavic, ed. Sue Brown & Adam Przepiórkowski, 1–30. Bloomington: Slavica.
Baerman, M. 2009. Case syncretism. In The Oxford Handbook of Case, ed. A. Mal’chukov & A. Spencer, 219–230. Oxford: Oxford University Press.
Beekes, R.S.P. 1995. Comparative Indo-European Linguistics: An Introduction. Amsterdam: John Benjamins.
Bybee, Joan, Revere Perkins, & William Pagliuca. 1994. The Evolution of Grammar: Tense, Aspect, and Modality in the Languages of the World. Chicago: University of Chicago Press.
Clackson, James. 2007. Indo-European Linguistics: An Introduction. Cambridge: Cambridge University Press.
Clackson, James, & Geoffrey Horrocks. 2007. The Blackwell History of the Latin Language. Malden, MA: Blackwell Publishing.
Croft, William. 2000. Explaining Language Change: An Evolutionary Approach. Harlow: Pearson Education Limited.
Davies, Mark. 2008–. The Corpus of Contemporary American English (COCA): 450 million words, 1990-present. Available on-line at <http://corpus.byu.edu/coca>.
Dixon, R.M.W. 1977. A Grammar of Yidiɲ. Cambridge: Cambridge University Press.
Givón, Talmy. 1979. On Understanding Grammar. New York: Academic Press.
Givón, Talmy. 1985. Iconicity, isomorphism, and non-arbitrary coding in syntax. In Haiman (1985: 187–219).
Givón, Talmy. 1995. Functionalism and Grammar. Amsterdam: John Benjamins.
Givón, Talmy. 2001. Syntax: An Introduction, v. 2. Amsterdam: John Benjamins.
Gonda, Jan. 1956. The Character of the Indo-European Moods with special regard to Greek and Sanskrit. Wiesbaden: Otto Harrassowitz.
Greenough, G.B. et al. eds. 1903. Allen and Greenough’s New Latin Grammar. Boston: Ginn & Company.
Haiman, John. 1983. Iconic and Economic Motivation. Language 59: 781–819.
Haiman, John (ed.). 1985. Iconicity in Syntax. Amsterdam: John Benjamins.
Jakobson, Roman. 1984. Contribution to the General Theory of Case: General meanings of the Russian cases. English translation in Russian and Slavic Grammar: Studies 1931–1981, ed. L. Waugh & M. Halle, 59–103. Berlin: Mouton. (Original work published 1936 as Beitrag zur allgemeinen Kasuslehre: Gesamtbedeutungen der russischen Kasus, Travaux du cercle linguistique de Prague 6; 240–288).
Jakobson, Roman. 1984. Morphological Observations on Slavic Declension (The Structure of Russian Case Forms). English translation in Russian and Slavic Grammar: Studies 1931–1981, ed. L. Waugh & M. Halle, 105–133. Berlin: Mouton. (Original work published 1958 as Morfologicheskie nabliudeniia nad slavianskim skloneniem. S-Gravenhage: Mouton & Co.).
Janda, Laura A., & Steven J. Clancy. 2002. The Case Book for Russian. Bloomington: Slavica.
Jing-Schmidt, Zhuo, & Vsevolod Kapatsinski. 2012. The apprehensive: Fear as endophoric evidence and its pragmatics in English, Mandarin, and Russian. Journal of Pragmatics 44: 346–373.
Kagan, Olga. 2010. Genitive objects, existence and individuation. Russian Linguistics 34: 17–39.
Kagan, Olga. 2013. Semantics of Genitive Objects in Russian. Dordrecht: Springer.
Kiparsky, Paul. 1968. Tense and Mood in Indo-European Syntax. Foundations of Language 4: 30–57.
Kirsner, Robert S. 1985. Iconicity and grammatical meaning. In Haiman (1985: 249–270).
Kitis, Eliza. 2009. From motion to emotion to interpersonal function: The case of fear predicates. In Language and Social Cognition: Expression of the Social Mind, ed. Hanna Pishwa, 433–454. Berlin: Mouton de Gruyter.
Kuryłowicz, Jerzy. 1927. Injonctif et subjonctif dans les Gāthās de l’Avesta. Rocznik Orientalistyczny 3: 164–179.
Lakey, Holly. 2012. Lest we fear the competition: Complement clause constructions after verbs of fearing in English. Unpublished manuscript.
Langacker, Ronald W. 1993. Reference-point Constructions. Cognitive Linguistics 4: 1–38.
Langslow, David (ed.). 2009. Jacob Wackernagel Lectures on Syntax with special reference to Greek, Latin, and Germanic. New York: Oxford University Press.
Lehmann, Winfred P. 1974. Proto-Indo-European Syntax. Austin: University of Texas Press.
“lest, conj.”. OED Online. December 2013. Oxford University Press. 16 January 2014. <http://www.oed.com/view/Entry/107487?redirectedFrom=lest>.
Lichtenberk, Frantisek. 1995. Apprehensional Epistemics. In Modality in Grammar and Discourse, ed. Joan Bybee & Suzanne Fleischman, 293–327. Amsterdam: John Benjamins.
Lunt, Horace G. 2001. Old Church Slavonic Grammar. 7th edition. Berlin: Mouton de Gruyter.
MacDonell, Arthur Anthony. 1927. A Sanskrit Grammar for Students. Third edition. Oxford: Oxford University Press.
Mathiassen, Terje. 1996. A short grammar of Lithuanian. Columbus: Slavica.
Mathiassen, Terje. 1997. A short grammar of Latvian. Columbus: Slavica.
Meier-Brügger, Michael. 2003. Indo-European Linguistics. Berlin: Walter de Gruyter.
Migdalski, Krzysztof. 2006. The Syntax of Compound Tenses in Slavic. Utrecht: LOT Publishing.
Morwood, James. 1999. A Latin grammar. Oxford: Oxford University Press.
Nau, Nicole. 1998. Latvian. München: Lincom Europa.
Neidle, Carol. 1988. The Role of Case in Russian Syntax. Dordrecht: Kluwer Academic Publishers.
Palmer, L.R. 1954. The Latin Language. London: Faber and Faber Limited.
Palmer, F.R. 1986. Mood and Modality. Cambridge: Cambridge University Press.
Partee, Barbara, & Vladimir Borschev. 2004. The semantics of Russian genitive of negation: The nature and role of perspectival structure. In Proceedings of the 14th Semantics and Linguistic Theory Conference, ed. Robert B. Young, 212–234. Ithaca: CLC Publications.
Pokorny, Julius. 1959. Indogermanisches etymologisches Wörterbuch. Bern: Francke.
Renou, Louis. 1928. Les forms dites d’injonctif dans le Rgveda. In Études de Linguistique, offertes par quelques amis à Emile Benveniste, 63–80. Paris: P. Geuthner.
Rozin, Paul, & Edward B. Royzman. 2001. Negativity Bias, Negativity Dominance, and Contagion. Personality and Social Psychology Review 5: 296–320.
Schmalstieg, William. 1983. An Introduction to Old Church Slavic. Columbus: Slavica.
Schwyzer, Eduard. 1950. Griechische Grammatik ii, Syntax und Syntaktische Stilistik. München: C.H. Beck.
Smyth, Herbert Weir. 1920. A Greek Grammar for Colleges. New York: American Book Co.
Timberlake, Alan. 1975. Hierarchies in the Genitive of Negation. The Slavic and East European Journal 19: 123–138.
Timberlake, Alan. 2004. A Reference Grammar of Russian. Cambridge: Cambridge University Press.
Watkins, Calvert. 1967. Remarks on the Genitive. Den Hague: Mouton.
Whitney, D.W. 1889. Sanskrit Grammar, including both the classical language, and the older dialects, of Veda and Brahmana. Boston: Ginn & Co.
van der Wurff, Wim. 1998. On Expletive Negation with Adversative Predicates in the History of English. In Negation in the History of English, ed. Ingrid Tieken-Boon van Ostade, Gunnel Tottie, & Wim van der Wurff, 295–327. Berlin: Mouton de Gruyter.
Zeijlstra, Hedzer Hugo. 2004. Sentential Negation and Negative Concord. (Doctoral dissertation). Retrieved on-line 4 May 2012 from <http://www.lotpublications.nl/publish/articles/001108/bookpart.pdf>.
I would like to thank the anonymous reviewers for their assistance. I would also like to thank Cynthia Vakareliyska, Vsevolod Kapatsinski, and Anna Mikhaylova for their helpful comments. All usual disclaimers apply.
In Indo-European languages, both *ne and *nē are well attested. See Pokorny 1959: 756–758. It is possible that *nē arose by monosyllabic lengthening with stressed negation. Thanks to the anonymous reviewer for pointing this out.
It is important to note that this article is concerned with case markings in specific languages, not with more universal notions of case in language in general.
The use of Russian here is as an example of the condition in Slavic languages. The existing Baltic (Lithuanian and Latvian) and Slavic (Russian, Ukrainian, Slovene, Slovak, Czech, Croatian, Serbian, Bosnian, Polish, Belarusian) languages that retain their morphological case systems all generally use the genitive case for the object of the verb “fear.” While the verb “fear” is often reflexive, meaning that its object cannot take the accusative case, there are other case endings that are used with reflexive verbs, e.g., the dative case. However, as fear verbs index this movement away from the fear stimulus or SOURCE, the genitive is the most appropriate case ending as the argument here sets out to show.
In the case of these subsets of Irrealis Genitive verbs that take the genitive of negation with a non-negative verb, when they are in turn negated, this absence of commitment to existence is no longer relevant, and therefore the irrealis conceptualization is cancelled out by the negation.
The injunctive had a range of functions and was not limited to prohibitions, although that is the usage that is discussed in the current analysis. For more discussion of the functions of the injunctive in IE languages, see Kuryɫowicz 1927, Renou 1928, Gonda 1956, and Kiparsky 1968.
Lithuanian behaves similarly in the use of a negative prefix ne-. See example (12) in section 4.2.2. Additionally, Latvian can also make use of the ne- prefix as well in fear constructions, although ne- + subjunctive mood is optional in Latvian, as the complementizer ka and the indicative mood can be used, as well.
See Croft’s (2000) discussion of cryptanalysis, in which covert semantic negativity leads to the overt expression of negation within utterances. A more traditional analysis, that these modal clauses arose from old negative optative constructions, holds that nē negates the main clause verb (e.g., “May he not die” → “I fear” + “may he not die” with nē becoming the complementizer between clauses). However, an optative source is not the only possible ancestral construction. A jussive source is also possible. Regardless of the historical source of fear constructions, the point to note here is that, synchronically, the negation is semantically driven.
The transcription is that of Kitis (2009: 421) with insertion of diacritics according to the Loeb edition by C.F. Smith (1959: 364).
However, French seems to be unique among the major Romance languages in having preserved this construction productively. Again, constructions with a complementizer with negative semantic content occur in many languages, as the examples here, and in section 4 with lest, illustrate. Their negative semantics make them suited for expressions of avoidance.
Fear lexemes are not the only ones to trigger the use of the expletive ne; it also occurs with empêcher ‘prevent,’ éviter ‘avoid,’ and in negative constructions with douter ‘doubt,’ and nier ‘deny.’ Interestingly, Old English also had an expletive ne which occurred in very similar contexts. Old and Middle English verbs meaning ‘doubt,’ ‘deny,’ ‘dread,’ ‘forsake,’ ‘hesitate,’ and ‘refuse’ could occur with this expletive negator. For example:
þan I haue no doute þat it ne schal wel kun telle þee of hem
then I have no doubt that it neg will fully be-able tell you of them
“Then I have no doubt that it will be able to tell you all about them”
(Cloud of Unknowing 92.6–7, cited after van der Wurff 1998: 295).
As Bybee et al. (1994: 181) states, “modality is the conceptual domain and mood is its inflectional expression.”
The subjunctive with the second or third persons would, then, indicate a request, which is still an exercise of the speaker’s will.
Both moods are continued in Vedic Sanskrit and Ancient Greek. I would like to thank an anonymous reviewer for pointing this out. In Greek, the situation is similar to the one outlined above for PIE, although the optative can replace the indicative or subjunctive mood in dependent clauses after main clause verbs in the past tense or historic present (e.g., in indirect speech and questions for the indicative, and fear and purpose clauses for the subjunctive).
For a diachronic discussion of modality and extension, see chapter six of Bybee et al. 1994. The authors argue that the disparate uses of a certain mood often seem “meaningless,” but, in actuality, extension from meaningful contexts (or harmonic ones) over time has led to its use in less semantically appropriate (non-harmonic) instances (213). The mood/modal element’s use in a subordinate clause arises from a harmony with the meaning of a main clause verb. The use of the optative or subjunctive in a main clause was extended to complement clauses with harmonic meaning, and later to non-harmonic contexts. In the case of fear constructions, the use of an irrealis mood arises out of this original harmony and is likely older than its use in non-harmonic contexts. Its use in fear constructions is rooted in the original notions of desire (or negated desire for fear), and these notions involve intention, which predates the more general notion of future.
Similar constructions have been labeled “apprehensives” or “apprehensive modality” constructions after Dixon (1977). This label may be more useful than “epistemic modality,” because it encompasses both the speaker-centered notion of epistemic modality and the apprehension brought on by the proposition. See Lichtenberk (1995) for a study of apprehensive modality.
For example, some fear constructions do not express a literal fear toward the proposition itself, but instead indicate a social fear by signaling that the listener may not like the content of the upcoming proposition. An example of this use from the Corpus of Contemporary American English is “I’m afraid we’re out of time” (ABC’s Nightline, 01/22/1991). This pragmatic function could be called a stance marker indicating an affective stance toward the message content. Regardless of labels, the function of mood and modality in fear constructions is to indicate the speaker’s conceptual distance to the content in the subordinate clause, whether the fear is of a possible negative event or of offending the listener.
The choice of subjunctive mood may be a result of the collapse of the subjunctive and optative moods in many daughter languages.
In Latvian the use of the indicative is also possible, and the ne- prefix on the verb is not used with the indicative (see section 3.2.1 above).
Although this paper concentrates on IE languages, other languages of the world share some of the same aspects of grammar to discuss fear. For example, several languages have a specialized complementizer like English’s lest that is used with fear constructions (see Lichtenberk 1995 for Austronesian languages).
The rise of modal auxiliaries in English has led to their preference over the use of subjunctive mood constructions, especially in contemporary American English. It could be argued that these auxiliaries have taken on the function of indicating distance as well, as they can be used to indicate the speaker’s stance towards the proposition, as well as indicating possibility/probability. See Lakey 2012 for a discussion of the development of complex English fear constructions.
See Croft’s (2000) discussion of cryptanalysis, in which covert semantic negativity leads to the overt expression of negation within utterances. A more traditional analysis, that these modal clauses arose from old negative optative constructions, holds that nē negates the main clause verb (e.g., “May he not die” → “I fear” + “may he not die” with nē becoming the complementizer between clauses). However, an optative source is not the only possible ancestral construction. A jussive source is also possible. Regardless of the historical source of fear constructions, the point to note here is that, synchronically, the negation is semantically driven.