Lietuvių kalba eISSN: 1822-525X
2022, vol. 17, pp. 49–60 DOI: https://doi.org/10.15388/LK.2022.4

The prehistory of kláusti, klausýti, and their related forms revisited

Yoko Yamazaki
Department for Slavic and Baltic Studies, Finnish, Dutch and German, Stockholm University
Email: yoko.yamazaki@balt.su.se
ORCID: https://orcid.org/0000-0002-0209-9599

Abstract. The Baltic verbs, Lith. kláusti ‘to ask questions’, klausýti ‘to listen’, Latv. klàust ‘to ask questions’ and klàusît / klaũsît ‘to listen’ exhibit intriguing morphological and accentological distributions. For a long time, this has interested linguists as shown by numerous previous studies starting with Schulze (1904). Despite several proposals put forward so far, researchers have not yet agreed on a historical interpretation of the acute tone of Lith. kláusti, or on the historical relationship of kláusti to its cognates not only in Balto-Slavic but also in Indo-European languages. In this paper, all the relevant data and hypotheses available to the current author will be examined again, and a more probable historical interpretation will be sought.

Keywords: Balto-Slavic historical morphology, Indo-European desideratives, iteratives, Baltic transitive ia-presents, Baltic intransitive inchoatives.

Dar kartą apie kláusti, klausýti ir jiems giminingų formų priešistorę

Santrauka. Baltų kalbų veiksmažodžiai, liet. kláusti, klausýti, latv. klàust ir klàusît / klaũsît pasižymi įdomiais morfologiniais ir akcentologiniais požymiais. Ilgą laiką tai domino kalbininkus, kaip rodo daugybė ankstesnių tyrimų, pradedant Schulze (1904). Nepaisant keleto iki šiol pateiktų pasiūlymų, tyrėjai dar nesutarė dėl istorinės kláusti akūtinio kirčiavimo interpretacijos ar dėl veiksmažodžio kláusti santykio su giminingais žodžiais ne tik baltų slavų, bet ir indoeuropiečių kalbose. Šiame straipsnyje bus dar kartą išnagrinėti visi autorei prieinami svarbūs duomenys ir hipotezės bei ieškoma labiau tikėtino istorinio aiškinimo.

Raktažodžiai: Baltų ir slavų kalbų istorinė morfologija, indoeuropiečių desideratyvai, iteratyvai, baltų tranzityvinės ia-kamieno esamojo laiko formos, baltų intranzityviniai inchoatyvai.

Received: 15/09/2022. Accepted: 27/12/2022
Copyright © 2022 Yoko Yamazaki. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

1. The problems surrounding Lith. kláusti ‘to ask questions’, klausýti ‘to hear’ and their cognates

Lith. kláusti ‘to ask questions’, klausýti ‘to hear’ and their corresponding words in Baltic (Latv. klàust, -šu; klàusît / klaũsît, -u, OPru. klausiton ‘to listen’) are well-known cognates of the Indo-European verbs for ‘to hear, listen’. They are exemplified by Skt. √śru- ‘to hear’, pres. (V) śṛṇóti, √śruṣ-, pres.-inj. 3pl. śróṣan (I. śróśati) ‘to listen, obey’; Gk. root-aor. impr. 2pl. κλῦτε ‘hear!’, OCS slušati, -ajǫ (variant slъšati), Ru. slúšat’, SCr. slȕšati < PS *slűšati (AP a), -ajǫ ‘to listen’, among others. The rich attestation of the cognates (cf. EWAía II 672, LIV² 334–336, LEW 265–266, Kroonen 2013, 232), however, presents a complicated picture, as suggested by the reconstruction of two similar verbal roots by LIV², i.e., *ḱleu- ‘to hear’ and *ḱleus- ‘to hear / listen (to)’. It appears that the complexity can be sorted out according to the following two parameters:

i) with or without the (root-) final laryngeal;
ii) with or without the stem-final *-s.

Surely, the tonal contrast between the acute tone (Lith. kláusti ‘to ask questions’, Latv. (variant) klaũsît, -u ‘to hear’) and the circumflex tone (Lith. klausýti klaũso, Latv. klàusît -u ‘to hear’, klàust, -šu ‘to ask questions’) constitutes a part of the complexity surrounding the verbal root *ḱleu- ‘to hear’ or *ḱleus- ‘to hear / listen (to)’. This tonal contrast of the Baltic forms has interested historical linguists for a long time, and several solutions have been proposed. In what follows, those proposals will be examined in a broader context of the variations that the Indo-European cognates exhibit. First, we will look into the variations observed among the cognates according to the aforementioned two parameters.

2. The variations of the cognates

As anticipated in § 1 above, the cognates exhibit variations. There may be various possible ways to sort those variations, but as a starting point of our discussion in this paper, we will attempt to sort them depending on the two parameters mentioned above, i.e., i) whether they have the root-final laryngeal; and ii) whether they have the stem-final *-s. The resulting classification is the following three groups:

1) Aniṭ *ḱleu- ‘to hear’ [without laryngeal, without *-s]

• Skt. √śru- ‘to hear’, pres. (V) śṛṇóti, root-aor. aśrot, fut. śroṣyāmi, śroṣyáti (B+), causative śraváyati, śrāváyati (< *ḱlo.u-e.ye/o-, note Brugmann’s Law).

• Gk. pres. act. κλέω ‘to celebrate’ / mid. κλέομαι < *ḱleu-e-(tor) (Vine 1998, 76), root-aor. impr. 2pl. κλῦτε ‘hear!’ with the replacement of older *-eu- with *-ū-, cf. the older *ḱléutē̆ with the root vocalism of *-eu- as attested in Ved. śrótā (cf. LIV² 334³).

• Activized medio-passives: OCS sluti, slovǫ ‘to be considered as, to be called’ < *ḱleu-e/o-(tor) (cf. Vine 1998, 76); Latv. sluvêt, sluv ‘to be known’ (possibly the original medio-passive inflexion shifted to the stative class in *-ē- common in Balto-Slavic, cf. LIV² 335⁶)

• Stative formations in *-ē-: Av. sruiiē ‘is famous’, Lat. clueō, -ēre ‘to be called, regarded as’.

2) Aniṭ *ḱleus- ‘to listen’ [without laryngeal, with *-s]

• Iterative: Lith. klausýti, klaũso, klaũsė, Latv. klàusît / klaũsît, -u ‘to listen’ < PIE *ḱlous-eye/o- ¹

• Inchoative: Lith. pa-klùsti, -klū̃sta / -klū̃sa, -klùso, Latv. klust, -stu ‘to obey’; OPru. poklusman ‘obedient’

• Stative: Lith. klusė́ti, klùsi, Latv. klusêt, -u ‘to be obedient’ as if from *ḱlus-eh₁-, pres. *ḱlus-i- (the semantics may speak for its derivation from the inchoative Lith. pa-klùsti ‘to obey’, Latv. klust ‘to become silent’)

• OCS slušati, -ajǫ (variant slъšati), Ru. slúšat’, SCr. slȕšati < PS *slűšati (AP a), -ajǫ ‘to listen’ < inf. *ḱlous-eh₁ -ti, pres. *ḱlous-eh₁-ye/o-; it is possibly a denominative from the verbal noun sluxъ ‘Gehör, Ruf’ (< PS *slȗxъ ‘hearing, rumour’, cf. LIV² 336¹). Verbal nouns are also attested in Ved. śruṣtí- ‘listening, obedience (verbal noun)’, Ave. s(ə)raoša- ‘obedience’, ON hlust n. ‘Gehör’, OE hlyst (m./f.) ‘hearing’. As shown in the variant slъšati, the root vowel did not necessarily develop from a Proto-Slavic long vowel, and the acute -ű- in PS *slűšati, which looks as though developed from a long vowel, might be analogically adopted from PS *sly̋šati (AP a) ‘to hear’.

• OPru. klausiton, pres. 1pl. klausēmai, VN poklausijsnan, impr. 2pl. klausieiti ‘to listen’ < inf. *klausēt, pres. *klausē-, corresponding to the Slavic inflexion exemplified by OCS slušati, -ajǫ (variant slъšati), according to Schmid 1963, 13ff.

• OHG hlosēn, SwG. lose ‘to listen’ < PG *hlusēn- < PIE *ḱlus-eh₁-ye/o- (Kroonen 2013: 232); Jasanoff (1987, 98) argues that the fact that OHG hlosēn ‘to listen’ belongs to the III-rd weak class may indicate that it “may have developed from an athematic present or perfect middle along the lines set forth in my Stative and Middle in Indo-European, ch. 3.²” In this case, OCS slyšati ‘to hear’, pres. slyši- belongs to the same stative class, although the Slavic form points to a root-final laryngeal.

• OE hlosnian, hlysnan < PGmc. *hlus-ni/a- ‘to listen’ < Northern IE *ḱlu-n-s-e/o- (Villanueva Svensson 2014, 245²³, Gorbachov 2007, 88f.)

• Skt.√śruṣ-, pres.-inj. 3pl. śróṣan, impr. 3pl. śróṣantu (I. śróśati), pres.-mid. ptpl. śróṣamāṇa- ‘to listen, obey’ < PIE *ḱléus-e/o-; also TochB 3sg. klyauṣäm, mid. klyauṣtär, TochA klyoṣtär (thematic present) support a PIE present stem *ḱléus-e/o-. However, they may not be descendants of the desiderative / future stem in *-s-. Jasanoff (1987, 99) suggests that these forms are possibly derived from the IE subjunctives of the s-aorist, adducing a Vedic hapax si-imperative 2sg. śróṣi ‘hear!’, which was probably formed from the s-aorist subjunctive stem (also Narten 1964, 45–49). Therefore, it should be noted that some sigmatic variants are not related to the desiderative / future stem.

3) Seṭ *ḱleuHs- ‘to hear’ [with laryngeal, with *-s]

• Skt. desiderative śúśrūṣate formed to the root √śru- ‘to hear’.

• OCS slyšati ‘to hear’, 1sg. slyšǫ, 2sg. slyšiši < PS *sly̋šati (AP a), *sly̋ši- < *slyxěti < *ḱlū-ṣ- (Stang 1966, 92) < *ḱluHs-eh₁-ti, *ḱluHs-?; Jasanoff (1987, 98ff.) and Klingenschmitt (2008, 206) treat this verb as one of the stative verbs in *-ē- (< *-eh₁-) with Narten ablaut *ḱlḗus- / *ḱléus-. This matter will be discussed below in § 3.

• Lith. kláusti, kláusia, kláusė ‘to ask, inquire (< *want to hear)’ < as if *ḱlóuHs-ye/o-, but the circumflex tone of Latv. klàust, -šu points to an aniṭ root / stem.

It may be noteworthy that there are no reflexes of seṭ root without *-s (**ḱleuH-) attested. It has been suggested by previous studies (Schulze 1904, 1434–1442, Gotō 1987, 317, Jasanoff 2003, 132ff. among others) that the variants with *-s have been introduced by the desiderative / future suffix *-(H)s-. Furthermore, Jasanoff (2003, 134) argues that the laryngeal in the suffix must have been an *-h₁-, which can be detected in Vedic desideratives and futures to roots endings in a resonant like kariṣyánt- ‘about to do’, cíkīrṣati ‘desires to do’, Gk. τενέω ‘I will stretch’ (< *-esō < *-h₁s-e/o-). In this view, the absence of laryngeal reflexes in the forms like Skt. vakṣyánt- ‘about to say’, Skt. vívṛṣati ‘desires to turn’ and Gk. γράψω ‘I will write’ is attributed to a laryngeal loss between obstruents already in PIE.

The absence of the seṭ variants without *-s (**ḱleuh₁-) arguably strengthens this view. It indicates that the laryngeal could not have occurred alone without *-s- in this extended root (*ḱleu-h₁s-), being a part of the desiderative / future morpheme *-h₁s-, while *-h₁- could be lost in specific environments through laryngeal loss rules. For example, the iterative stem *ḱlóu-[h₁]s-eye/o- and the nominal stem *ḱlou[h₁]s- (> PS *slȗxъ ‘hearing, rumour’), which served as the basis of the denominative *ḱlous-eh₁-ti (> PS *slűšati (AP a), -ajǫ ‘to listen’ ), satisfy the structural description of a laryngeal loss rule called “the Saussure Effect”.³ In this way, it is easily understandable that the aniṭ variants with *-s (i.e.,*ḱleus-) could be introduced in the proto language, while a seṭ variant without *-s (**ḱleuh₁-) could have much fewer chances to occur.

3. Previous proposals on the prehistory of Lith. kláusti ‘to ask questions’ and klausýti ‘to hear’

Schulze (1904, 1434ff.) was one of the people in the early era of Indo-European studies, among others⁴, who were aware of the historical relationship between the Indic future in -sya- and the desiderative. Schulze noted the Sanskrit data attesting the shortening (or the loss of -i-) in the future forms formed to monosyllabic roots ending in resonants, e.g., maniṣyá- ‘will think’ → 3sg. maṃsyáte to √man-; staviṣya- ‘will praise’ (1sg. staviṣyā́mi) → stoṣya- (3sg. stoṣyáti) to √stu-. This indicates that these future forms, although configured to what we now call “aniṭ roots”, behaved earlier as though formed to seṭ roots. Schulze saw a parallelism in the desiderative forms, which also behave as though they were formed to seṭ roots, even when they are actually formed to aniṭ roots. Compare the lengthening of the root vowel in desiderative śúśrūṣate to aniṭ √śru- ‘to hear’ and desiderative búbūṣate to seṭ √bhū ‘to be’.⁵ This confirmed for him the close relationship between the future and the desiderative. He further suggested that in the same fashion that maniṣyá- ‘will think’ was shortened to 3sg. maṃsyáte, the future 1sg. śroṣyāmi to √śru- ‘to hear’ must also have been a result of the loss of -i- in an earlier *śraviṣyā́mi. Schulze equated this reconstructed future *śraviṣyā́mi with Lith. kláusiu ‘I ask questions’ (*kleuәs-). On the other hand, Shulze analysed the circumflexed Lith. klaus-aũ ‘I listen’ to be formed to the root *ḱleus- (aniṭ and sigmatic), equating it with Skt. √śruṣ- (śróṣamāṇa-, śruṣtí-) and Gmc. *hlus- (OS hlust, OE hlyst, OHG hlosēn, OE hlosnian). Schulze left the historical origin of the aniṭ sigmatic variant unexplained. However, as discussed above (§ 2), it could be introduced through the operation of a laryngeal loss rule in a relevant stem or some reflexes of it (i.e., Skt. śroṣi ‘hear!’, TochB 3sg. klyauṣäm, mid. klyauṣtär, TochA klyoṣtär) had a totally different origin, i.e., the s-aorist subjunctive (see above for the discussions of Narten 1964 and Jasanoff 1987). Nonetheless, of course, except for these Indo-Iranian and Tocharian forms whose origin has been suggested to be another, the desiderative reading of the extended root *ḱleu-(h₁)s- seems to have been generally accepted, as can be seen in LEW 265, EWAía II 672, García-Ramón (1994, 58–59), Derksen (1996, 312) among others.

Schmid (1962) agrees with Schulze that kláusti originated from a desiderative formation with the suffix -s-. He believes that the circumflex tone of the root vowel -aũ- (cf. 3sg./pl. klaũso ‘to listen’) should be regular in Baltic ia-presents (< IE *ye/o-presents), and therefore the acute tone of kláusti, -ia ‘to ask questions’ cannot be a regular phonological outcome. His investigation starts from the Baltic root *klaus-, whose meaning ‘to hear’ is attested in all the Baltic languages. An original ia-present paradigm, i.e., inf. *klaũsti, pres. 1sg. *klausiù, 3sg./pl. *klaũsia ‘to ask questions’, is reconstructed for Proto-Baltic, which would have a preterit 3sg./pl. *klaũsē ‘asked questions’ homonymic with the preterit of klausýti ‘to listen’. Thus, Schmid proposed that the acute tone was introduced to *klaũsti *klaũsia, *klaũsē ‘to ask questions’ in order to differentiate it from *klaũsyti, *klaũsia, *klaũsē ‘to listen’.

A similar line of approach is presented by Otrębski (1963). His proposal is a paradigm split. Otrębski assumes that Proto-Baltic inherited one single paradigm, i.e., inf. *klaũsīti, pres. 3sg.(/pl.) *klaũsi ‘to listen’ with the circumflex tone, from which two new paradigms split out. One of them is inf. *klausīti, pres. *klaũsā ‘to listen’ with the circumflex tone and new iterative present stem in *-ā-, and the other is inf. *klaũsti, pres. *klaũsi̯a with the new infinitive stem and the new meaning ‘to ask questions’. The new infinitive stem received the secondary acute tone, which spread to the whole paradigm of the verb for ‘to ask questions’ in Lithuanian (inf. kláusti, pres. kláusia), while the circumflex tone has been maintained in Latvian (klàust, klàušu ‘to ask questions’). This implies that this paradigm split process was in East Baltic.

However, not everyone was convinced by those proposals above. Bammesberger (1991) casts a spotlight on a class of ya-present verbs with the acute -áu- root vocalism: e.g., láužti, láužia, láužė ‘to berak’ (IE *leuǵ-⁶) láukti, láukia, láukė ‘to wait’ (IE *leuk- ‘to shine’), where circumflex -aũ- might be expected. In order to account for the acute tone certainly unexpected for at least some of them, a denominative derivation is proposed: a vṛddhi formation of *klaũsa- ‘hearing’ → *kláus-a- ‘interrogating’, to which a verbal stem *kláusia- ‘to ask questions’ was formed. Lith. klausà (acc. sg. klaũsą) ‘ability to hear’ attests the non-acute nominal stem *klaũsa- ‘hearing’ < IE *ḱlou-s-. Therefore, at least the basis for a vṛddhi formation *ḱlṓu-s-o- existed in Baltic. Likewise, for láukti, a nominal stem *láukas ‘*what belongs to the field, a guard, watch’ (< *lṓuk-o-), a vṛddhi formation of *louk-o- (> Lith. laũkas ‘field’) is postulated. However, the semantics of the proposed vṛddhi does not seem to be supported by the acute variant Latv. laũks ‘field’. Semantically, it would be more attractive to derive it from a causative-iterative *láukyti (> OPru. laukīt ‘look for’), with a metatonical acute root since métatonie rude often takes place in causative-iteratives (Derksen 1996, 310ff.). Nonetheless, even Bammesberger’s (1991) proposal has not convinced everyone. Smoczyński (2001, 159–160) finds Bammesberger’s (1991) proposal unconvincing because of the absence of attested parallel developments. Ostrowski (2001, 181) presents his scepticism to the proposal, stating that the type of nominal derivation advanced by Bammesberger (1991) for the adduced verbs, including for *ḱleu-s-, does not seem to be well established and, therefore, this solution does not seem fully convincing.

As an alternative to Bammesberger’s approach, Smoczyński (2001, 159–160) proposed that the secondary acute tone of kláusti, kláusia could have resulted from a syncope of *-i- in its preform *klaũsija (< *klous-eye/o-). The syncope could cause the compensatory lengthening of the diphthong *-aũ- in the root, and the resulting long diphthong received the acute tone as in kláusia. He adduces the acute tone of ia-present verbs occurring side by side with iteratives, e.g., láužti, láužiu ‘to break’ (~ iter. láužyti), spáusti, spáudžiu ‘to press’ (~ iter. spáudyti), as the parallel cases. Nevertheless, there seems to be a difference between the case of kláusti, kláusia and the case of spáusti, spáudžiu. While the present of both spáusti and spáudyti has the acute tone (spáudžia and spáudo, respectively), which allegedly is the result of the syncope in older *spaudija, the present of klausýti, i.e., klaũso (← *klaũsija), does not have the acute root.⁷ This probably means that factors other than the apocope of *-i- cause the acute tone of kláusti, kláusia.

Jasanoff (1987, 98ff.) and Klingenschmitt (2008, 206), as noted earlier, take a completely different approach. Jasanoff (1987) holds the view that OCS slyšati ‘to hear’, pres. slyši- (< PS *sly̋šati (AP a), *sly̋ši- < *slyxěti < *ḱlū-ṣ-, cf. Stang 1966, 92) is one of the stative verbs in *-ē- (< *-eh₁-) with Narten ablaut *ḱlḗus- / *ḱléus-. Klingenschmitt (id.) further envisages a development of PIE *ḱlḗus- / *ḱléus- to PS pres. *ślýši- (< PBS 3pl. *’ślūš-inti) through an earlier *’śl’auš- / ‘ślūš-. Since the Narten ablaut pattern of this verbal root is confirmed by Tocharian forms (A klyoṣ-, B klyauṣ(ä)- / klyauṣe- ‘to hear’ < PT *kl’aus’ä- / *kl’ausa- < *ḱlḗus-e/o-), this is accepted as a possible scenario, for example, in Villanueva Svensson (2014, 244ff.). However, it would also be possible to assume a more straightforward (or direct) development from *ḱluHs-eh₁-ti, *ḱluHs- (through PBS 3pl. pres. ślūš-inti) to PS *sly̋šati (AP a), *sly̋ši-.

As reviewed so far, various solutions and views have been put forward regarding the etymological background of Lith. kláusti ‘to ask questions’ and klausýti ‘to hear’; however, researchers have not yet agreed on a convincing conclusion concerning the prehistory of these verbs and their related Baltic forms. At least two points seem to be generally accepted. The first point is Schulze’s (1904) insight that the sigmatic variants (Baltic *klaus-, *kleus-) originated from an IE desiderative / future formation (*ḱleu-h₁s- ‘*want to hear’ > ‘listen [to]’). The second is that Lith. klausýti is an iterative formation, be it an inherited IE formation or a Baltic derivation. Nevertheless, researchers have not agreed on the prehistory of kláusti and klausýti.

4. Remaining questions

Through the quick review of the previous studies and the data (§§ 2, 3) emerge some questions remain from the previous studies. The most concerning question for the current author is whether PB *klaus-tē̆i, pres. *klaus-ya- underlying Lith. kláusti and Latv. klàust ‘to ask questions’ (as if < *ḱlouh₁s-ye/o-) can be directly equated with the Indic future (*ḱleuh₁s-ye/o-). Alternatively, in other words, whether this formation is so old that it is a direct descendant of a Proto-Indo-European formation. Schulze (1904) is without a doubt correct in relating these sigmatic variants to the IE desiderative / future formations. However, if a PIE form *ḱlouh₁s-ye/o- existed, the laryngeal would have been lost by the Saussure effect and the acute accent of Lith. kláusti cannot be the regular outcome. The absence of the palatalisation of /l’/ also indicates that the vocalism of kláusti is not precisely that of *ḱleuh₁s-ye/o- reconstructed for the Indic future (cf. Ostrowski 2001, 181). This may be relevant to the second question below.

Second, it is still not totally clear what sort of root vocalism and stem formation would be chosen in the individual desiderative / future formations in PIE. Jasanoff (2003, 135) suggests the possibility that the desiderative formations may correlate to their present formations; for instance, roots that made Narten-presents may have favoured Narten s-desideratives, while the roots that made reduplicated presents may have chosen reduplicated desideratives. Accordingly, at first glance, a desiderative stem *ḱlouh₁s-ye/o- appears to point to the existence of a present stem in *ḱlou-ye/o-. However, it is difficult to find any comparative support for the root *ḱleu- forming such a present stem. It would rather be more reasonable to assume that the o-grade-like root vocalism of kláusti has a more recent (in other words, not as old as Proto-Indo-European) prehistory. One might think of the iterative *ḱlouh₁s-eye/o- as its source. In my view, Schmid (1962), Otrębski (1963) and Bammesberger (1991) correctly assumed that the acute tone of kláusti was secondarily introduced in one way or another.

The failure of sibilantisation of the initial *ḱ- is considered to be due to the dissimilation effect of the desiderative morpheme *-h₁s- (see Stang 1966, 91–92). This, together with the Slavic cognates, indicates the antiquity of the desiderative stem *ḱleuh₁s-, which was probably reinterpreted as a root for “to listen to” in Balto-Slavic or possibly even earlier. On the one hand, Lith. klausýti / Latv. klàusît ‘to listen’ exhibits the regular outcome of the iterative *ḱlou[h₁]s-eye/o-, and therefore, the second question is how the acute tone of kláusti, on the other hand, has been introduced in the prehistory of the Baltic branch.

5. Proposal

Turning our eyes to Slavic, we find the following verbal attestations:

• denominative: OCS slušati, -ajǫ (variant slъšati), Ru. slúšat’, SCr. slȕšati < PS *slűšati (AP a), -ajǫ ‘to listen’ < inf. *ḱl(o)u[h₁]s-eh₁ -ti, pres. *ḱl(o)u[h₁]s-eh₁-ye/o-,

• possibly descendants of the ancient medio-passive: OCS sluti, slovǫ ‘to be called’ < PS *slut, *slȍvǫ ‘be called, be famed’ < *ḱleu-tei, *ḱleu-e/o-tor,

• stative: OCS slyšati ‘to hear’, 1sg. slyšǫ, 2sg. slyšiši < PS *sly̋šati (AP a; < *sly̋xěti), *sly̋ši- < PBS *s/klṵ̄sḛ̄ti, *s/klṵ̄si- < PIE *ḱlḗus- / *ḱléus- or *ḱluh₁s-.

Out of these three, the stative paradigm provides the unambiguous reflex of the acute root, continuing either Narten type *ḱlḗus- / *ḱléus- or the laryngeal in *ḱluh₁s-. However, the stative formation of this stem is not preserved in Baltic⁸, where one finds *girdētē̆i (Lith. girdė́ti, Latv. dzirdēt, OPru. kirdīt, kīrdimai ‘to hear’), in place of *klṵ̄sḛ̄tē̆i (as if < *ḱluh₁s-eh₁-ti). The reason remains unclear why this lexical replacement took place in Baltic. Nonetheless, it would be worthy to clarify how those words formed to the root *ḱleuh₁s-, including how the stative PBS *s/klṵ̄sḛ̄ti, *s/klṵ̄si-, were treated in Baltic. The answer to the question as to how the acute tone of Lith. kláusti has been introduced in Baltic might reside in the Baltic treatments of the family of *ḱleuh₁s-based words.

As already envisaged (cf. Bammesberger 1991, Ostrowski 2001, 181), Lith. kláusti / Latv. klàust must certainly be a secondary formation based on an inherited verbal form. However, that does not mean that kláusti has to be directly derived from klausýti (*ḱlou[h₁]s-eye/o-). In this paper, it is proposed that Lith. kláusti / Latv. klàust could be a transitive-terminative formation based on, or at least influenced by the lost Baltic descendant of the stative formation equivalent to OCS slyšati ‘to listen’, 1sg. slyšǫ, 2sg. slyšiši < PS *sly̋šati (AP a), *sly̋ši- < *ḱlṵ̄-ṣ- (cf. Stang 1966, 92) < late PBS *s/klṵ̄sḛ̄ti, *s/klṵ̄si-⁹ (< PIE *ḱluh₁s-eh₁-ti, *ḱluh₁s- or Narten ablaut *ḱlḗus- / *ḱléus-). The descendant of the late PBS *s/klṵ̄sḛ̄ti, *s/klṵ̄si- is unfortunately not attested in Baltic, probably because it was replaced by *girdḛ̄tē̆i (Lith. girdė́ti, Latv. dzirdēt, OPru. kirdīt, kīrdimai ‘to hear’) for a reason yet unknown. However, before the replacement took place, if late PBS *s/klṵ̄sḛ̄ti, *s/klṵ̄si- still existed in Proto-Baltic, it would have given rise to PB *klṵ̄sḛ̄tē̆i (pres. *klūsi-) ‘to hear, be hearing’. A transitive-terminative form to the same root could have adopted the acute tone from PB *klṵ̄s(i)- to give rise to *kla̰usia- (with acute root) when it was formed according to a common Baltic pattern of verbal formations.

The table below shows the common Baltic pairs of transitive verbs in ia-presents and ė-preterits and intransitive-inchoative verbs in either nasal presents or sta-presents (cf. Stang 1942, 124–125; Endzelīns 1951, 764–765). In addition, stative or intensive formations in -ėti based on the respective verbal roots in the table are quoted when available. The pattern is partially paralleled by Slavic verbs as supplemented with OCS forms.

Table 1. Baltic transitive ia-presents, intransitive-inchoatives and verbs in -ėti

tran. ia-pres. & ė-pret.	intr.-inch. pres.	Stative or intensive in -ėti
Lith. baũsti, -džia, -ė ‘to punish’	Lith. bùsti, buñda, bùdo OCS vъz-bъ(d)nǫti, -bъdъ ‘to wake up’	Lith. budė́ti, bùdi, budė́jo OCS bъděti, bъždǫ ‘to be awake’
Lith. délti, dẽlia, dė́lė ‘to make blunt’	Lith. dı̀lti, dỹla / dẽla, dı̀lė ‘to become blunt(er)’	Lith. delė́ti, dẽli, delė́jo ‘to be sick’
Lith. kélti, kẽlia, kė́lė Latv. cel̂t, ceļu, cêlu(ē) ‘to lift up’	Lith. kı̀lti, kỹla / kı̀lsta, kı̀lo Latv. cilt cilstu cilu ‘to rise up’
Lith. maũkti, maũkia, maũkė ‘to set free’	Lith. mùkti, muñka, mùko ‘to come off’ OCS (pro-)mъknǫti sę, -mъkъ ‘to spread’	OCS mъčati, mъčǫ (mъči-) ‘to throw back and forth’
Lith. láužti,láužia, láužė Latv. laûzt, -žu, -zu ‘to break (tr.)’	Lith. lū́žti, -žta, -žo ‘to break (intr.)’	Lith. lū́žėti, -ėja / -i, -ėjo ‘to break a little’
Lith. kláusti, kláusia, kláusė Latv. klàust, -šu ‘to ask questions’	Lith. paklùsti,-klū̃sta, -klùso Latv. klust, -stu, -u ‘to obey’	PB *klṵ̄sḛ̄tē̆i, pres. *klūs(i)- ‘to hear’

As can be seen from some empty slots under “stative or intensive in -ėti”, the stative (or intensive) formation in -ėti is not a regular part of this pattern. However, the pairs of transitive ia-presents and intransitive-inchoatives are prevalent in East Baltic. Therefore, it is not conceivable that transitive ia-present Lith. kláusti / Latv. klàust is directly derived from the stative *klṵ̄sḛ̄tē̆i, pres. *klūs(i)- or iterative *klausītē̆i. However, in this particular case, the ia-present verb Lith. kláusti / Latv. klàust ‘to ask questions’ seems semantically more closely related to iterative klausýti ‘to listen’ and stative PB *klṵ̄sḛ̄tē̆i ‘to hear’ rather than to its intransitive-inchoative counterpart, paklùsti ‘to obey’. The meaning of kláusti, ‘to ask questions’, does not directly correspond to the meaning of Lith. paklùsti ‘to obey’ / Latv. klust ‘to become silent’. Probably Lith. paklùsti / Latv. klust has gone through a semantic change ‘*to start to listen > to become silent > to obey’, probably starting from the iterative *klausītē̆i ‘to listen’. The non-acute root would also indicate that they were formed based on klausýti ‘to listen’, which also has the non-acute root (cf. 3sg./pl. pres. klaũso). For that matter, the attested stative formation Lith. klusė́ti / Latv. klusêt ‘to be obedient’ is a perfect match both semantically and accentologically with the non-acute Lith. paklùsti ‘to obey’ / Latv. klust ‘to become silent’. On the other hand, Lith. kláusti / Latv. klàust ‘to ask questions’ seems to have developed a new meaning on its own, starting from ‘to hear, listen’, i.e., ‘*want to hear [something particular] > ask questions’. The acute tone of Lith. kláusti indicates some influence from the lost base stative formation *klṵ̄sḛ̄tē̆i ‘to hear’.

Through the discussion above on the semantics and the tone of the root, it has been attempted to show that Proto-Baltic inherited the iterative *klausītē̆i and stative *klṵ̄sḛ̄tē̆i, and the intransitive-inchoative (*klustē̆i, pres. *klu-n-sa), transitive ia-present (*klausti, *klausia), and the new stative (*klusḛ̄tē̆i) are Baltic creations. While it is relatively straightforward that the intransitive-inchoative (*klustē̆i, pres. *klu-n-sa) was formed based on the iterative *klausītē̆i and the new stative (*klusḛ̄tē̆i) was further formed to the intransitive-inchoative (*klustē̆i, pres. *klu-n-sa), the transitive ia-present (*klausti, *klausia) could have been formed either to the iterative *klausītē̆i or to the stative *klṵ̄sḛ̄tē̆i. Since the iterative *klausītē̆i inherited the non-acute root, while the stative *klṵ̄sḛ̄tē̆i had the acute root, it is conceivable that the Baltic derivatives confused these different tones in their derivational processes or post-derivational analogical processes. This may explain why the Latvian iterative form itself has a tone variation with the acute root, i.e., klaũsît ‘to listen’ (beside klàusît); and why Baltic transitive ia-present *klausti, *klausia ‘to ask questions’ has adopted different tones in Lithuanian (kláusti) and Latvian (klàust).

6. Conclusions

This paper has attempted to answer the ongoing, long-standing question, namely, why the Baltic verbs for ‘to listen, hear, ask questions’ exhibit seemingly random tone variations. First, it was attempted to sort out the attestations across the Indo-European and Balto-Slavic languages. Through the sorting processes, it has been revealed that there are reflexes of aniṭ *ḱleu-, aniṭ *ḱleus-, and seṭ *ḱleuHs-,but no reflexes of seṭ **ḱleuH-. For Slavic, unambiguous reflex with the acute root, continuing either Narten ablaut *ḱlḗus- / *ḱléus- or the seṭ variant *ḱleuHs-, is the stative formation with the acute root PBS *s/klṵ̄s-ḛ̄-ti, *s/klṵ̄s(-i)-, descended by OCS slyšati ‘to listen’, pres. slyši- ‘to hear’.

Accepting the previous studies, it is assumed in this paper that the variants with *-s and the root-final laryngeal were introduced in Proto-Indo-European by the desiderative suffix *-h₁s-. The non-acute reflex of the iterative Lith. klausýti and Latv. klàusît is proposed to have been introduced through an old laryngeal loss in Proto-Indo-European, i.e., the Saussure effect in PIE *ḱlouh₁s-eye/o- > *ḱlous-eye/o-.

Following this argument, it has been hypothesised that Proto-Baltic probably inherited the iterative PB *klausītē̆i ‘to listen’ (< PIE *ḱlous-eye/o-) with the non-acute root and the stative PB *klṵ̄sḛ̄tē̆i ‘to hear’ (< PIE *ḱlḗus- / *ḱléus- or *ḱluh₁s-eh₁-) with the acute root. It was concluded that these tone variations in the root inherited in Proto-Baltic must have caused the random-looking tone variations in Lith. kláusti / Latv. klàust ‘to ask questions’ and Latv. klaũsît beside klàusît ‘to listen’.

References

Bammesberger, Alfred. 1991. Zur Vorgeschichte von lit. kláusti und mìršta. Historische Sprachforschung / Historical Linguistics 104 (2), 269–278.

van Beek, Lucien. 2011. The “Saussure Effect” in Greek: a reinterpretation of the evidence. Journal of Indo-European Studies 39 (1/2), 130–175.

Delbrück, Berthold. 1874. Das altindische Verbum. Halle: Waisenhaus.

Derksen, Rick. 1996. Metatony in Baltic. Amsterdam: Rodopi.

García Ramón, José Luis. 1994. Zur historischen Betrachtung der indogermanischen Aktionsarten und Aspektprobleme: idg. *neu̯(H)- ʻeine momentane Bewegung machen, (sich) einen Augenblick wenden, drehenʼ. Münchener Studien zur Sprachwissenschaft 54, 33–63.

Gorbachov, Yaroslav. 2007. Indo-European Origins of the Nasal Inchoative Class in Germanic, Baltic and Slavic. PhD thesis. Cambridge, MA: Harvard.

Gotō, Toshifumi. 1987. Die “I. Präsentklasse” im Vedischen. Untersuchung der vollstufigen thematischen Wurzelpräsentia. Wien: Österreichische Akademie der Wissenschaften.

EWAía – Manfred Mayrhofer, ed. 1996. Etymologisches Wörterbuch des Altindoarischen, vol. 2. Heidelberg: Carl Winter.

Jasanoff, Jay. 1978. Stative and middle in Indo-European. Innsbruck: Institut für Sprachwissenschaft.

Jasanoff, Jay. 1987. Some irregular imperatives in Tocharian. In Studies in Memory of Warren Cowgill (1929-1985), Calvert Watkins, ed. Berlin and New York: De Gruyter, 92–112.

Jasanoff, Jay. 2003. Hittite and the Indo-European verb. New York: Oxford University Press.

Klingenschmitt, Gert. 2008. Erbe und Neuerung bei Akzent und Ablaut in der litauischen Morphologie. Danguolė Mikulėnienė, Laima Grumadienė, Saulius Ambrazas (eds.) Kalbos istorijos ir dialektologijos problemos 2. Vilnius: Lietuvių kalbos institutas, 180–215.

Kroonen, Guus. 2013. Etymological Dictionary of Proto-Germanic. Leiden & Boston: Brill.

LEW – Ernst Fraenkel (1962-1965). Litauisches etymologisches Wörterbuch. Heidelberg: Carl Winter.

LIV² – Helmut Rix, et al. eds. 2001. Lexikon der indogermanischen Verben. Die Würzeln und ihre Primärstammbildungen. Unter Leitung von Helmut Rix bearbeitet von Martin J. Kümmel, Thomas Zehnder, Reiner Lipp, Brigitte Schimmer. 2nd ed. Wiesbaden: Ludwig Reichert.

Narten, Johanna. 1964. Die sigmatischen Aoriste im Veda. Wiesbaden: Otto Harrassowitz.

Nussbaum, Alan. 1997. The “Saussure Effect” in Latin and Italic. Lubotsky, A. (ed.), Sound Law and Analogy: Papers in honor of Robert S. P. Beekes on the occasion of his 60th birthday. Amsterdam and Atlanta: Rodopi, 181–203.

Olander, Thomas. 2015. Proto-Slavic inflectional morphology. Leiden & Boston: Brill.

Ostrowski, Norbert. 2001. Angeblich primäre Jotpräsentien im Litauischen. Historische Sprachforschung 114, 177–190.

Otrębski, Jan. 1963. Noch einmal über lit. Kláusti. Indogermanische Forschungen 68, 42–46.

Pronk, Tijmen. 2011. The Saussure effect in Indo-European Languages Other Than Greek. Journal of Indo-European Studies 39 (1/2), 176–193.

Rasmussen, Jens. 1989. Studien zur Morphophonemik der indogermanischen Grundsprache. Innsbruck: Institut für Sprachwissenschaft.

de Saussure, Ferdinand. 1905. D‘ὠμήλυσις à Τριπτóλεμος, remarques étymologiques. In Mélanges Nicole. Genève: W. Kündig & Fils. 503–514.

Schmid, Wolfgang P. 1962. “Lit. kláusti „fragen“ und das altind. Futurum” Indogermanische Forschungen 67(1), 1–15.

Schmid, Wolfgang P. 1963. Studien zum baltischen und indogermanischen Verbum, Wiesbaden: Otto Harrassowitz.

Smoczyński W. 2001. Zu den litauischen mit der Binnensilbensynkope zusammenhängenden akutierten Wortformen. Linguistica Baltica IX, 155–162.

Stang, Christian S. 1942. Das slavische und baltische Verbum. Oslo: Dybwad.

Stang, Christian S. 1966. Vergleichende Grammatik der baltischen Sprachen. Oslo: Universitetsforlaget.

Schulze, Wilhelm. 1904. Lit. kláusiu und das indogermanische Futurum. Sitzungsberichte der Preußischen Akademie der Wissenschaften, 1434–1442.

Villanueva Svensson, Miguel. 2014. Tone variation in the Baltic ia-presents. Indogermanische Forschungen 119(1), 227–249.

Vine, Brent. 1998. Aeolic ὄρπετον and Deverbative *-etó- in Greek and Indo-European. Innsbruck: Institut für Sprachwissenschaft.

Yamazaki, Yoko. 2009. The Saussure Effect in Lithuanian. Journal of Indo-European Studies 37, 430–461.

---