Various sources I've read (e.g., the Wikipedia entry on Klingon) say that Klingon has two moods: indicative and imperative. Maybe this is right. But what determines whether something is a mood? Does Klingon have an interrogative mood (*bItlhutlhtaH'a'*/Are you drinking?/)? What about an optative mood (*bIQuchjaj*/May you be happy/)? We know that we can create an irrealis with *net jalchugh; *is this a subjunctive mood? Or does a mood require an inflected verb? Do the qualification suffixes create a mood? Basically, how does one determine what moods a language actually has, and can we apply this to Klingon? -- SuStel http://trimboli.name
ghItlhpu' SuStel, jatlh:
Basically, how does one determine what moods a language actually has, and can we apply this to Klingon?
My understanding is that "mood" as linguistically defined refers to specifically morphological means of signalling modality (that is, the speaker's subjective attitude towards the action of the verb as it is, should be, or may be) on finite verbs, since lexical expressions of modality are presumably limitless in any given language. Also, I think moods in a morphological system of modality would need to be mutually exclusive, otherwise you'd start multiplying moods together and end up with a much more complicated system. A challenge in distinguishing what constitutes mood and what doesn't is that the analysis varies from linguist to linguist; some linguists happily treat evidentiality, for instance, as a subset of modality, while others prefer to keep the two distinct (that's my perspective, since evidentiality is more about an objective statement of one's evidence for an assertion about what is, as opposed to one's subjective attitude towards what is, what should be, or what may be). You might like to read Ferdinand de Haan's discussion on evidentiality and modality (begins on p.47 of the PDF): https://pdfs.semanticscholar.org/f433/1055c9d78e29d7024f7d966fe22e5c749eac.p... Now, some tedious pondering on my part as to what this means for Klingon. Take all that follows with a good-sized grain of salt. With the idea of clearly separating evidentiality from modality, I'd probably treat Klingon as having four distinct morphological moods: one realis (indicative), two deontic (imperative, optative -jaj), and one epistemic (interrogative -'a'). The other Type 9s are all basically involved in making a verb non-finite, either as subordinate clauses (-DI', -bogh, -chugh, -pa', -vIS, -meH) or nouns (-ghach, -wI'), and because non-finite, modality doesn't really apply. Whether one would want to include the Type 2 suffixes in a count of moods is a more open question. I would tend to say no, because they seem to be able to happily co-occur with the non-indicative moods as well. In Turkish, -mAlI- forms (e.g. gelmeliyim "I must come") are often treated as a distinct necessitative mood, and the difference there is that a verb cannot be necessitative and any other mood at the same time: you can't have an imperative necessitative, for instance. Whereas in Klingon, as far as I know we can quite happily say something like mejnISjaj "may he have to leave" (even if the circumstances in which one would say something like that would be rather narrow). The Type 6 qualification suffixes fall pretty neatly into the category of evidentiality (and for me therefore not mood, though again, depends on one's analysis). For type 3, I think that's more aspect, though not the classic forms seen in the Type 7 aspect suffixes. -choH corresponds pretty neatly to what linguists call inchoative aspect, which is not uncommon in various languages. Much rarer is the idea of resumption that -qa' shows, but the Google overlords indicate there's a morphological resumptive aspect "again, starting again" in Kiliwa, a language of Baja California in Mexico. (Tangentially, Kiliwa was the dissertation topic of one Mauricio Mixco, who completed his doctorate at UC Berkeley, in the 1970s, and under the supervision of Mary Haas. Coincidence? I doubt it.) Type 8 (-neS) seems clearly just an honorific, with no other real semantic function (a couple of strange KGT examples aside). Type 4 (-moH), as a valency-changing device, would usually be referred to as voice (cp. other valency-changing devices like passive, middle, antipassive, etc.). Finally, Type 5. Here things get kind of interesting. -lu' isn't passive as such; we know that. But like the English passive, and indeed the Klingon causative, it's a valency-changing operation, therefore voice. -laH is a bit different because it doesn't actually alter valency in modern Klingon, so it's hard to call it voice sensu stricto. However, I've just discovered that in Japanese grammar what's usually called the "potential voice" originates from a form not unlike the passive, and has two case-marked forms, one of which does alter valency: Active: Tomoko ga mizu o nomimasu "Tomoko drinks water" Passive: Mizu ga (Tomoko ni) nomaremasu "water is drunk (by Tomoko)" Potential: Tomoko ga mizu o nomemasu "Tomoko is able to drink water" Potential: Mizu ga (Tomoko ni) nomemasu "water is drinkable (by Tomoko)" It may be worthy of noting that English deverbal adjectives in -able are generally passive in nature too: drinkable is not "able to drink", but "able to be drunk". Perhaps an insight (within the game, at least) into the historical origin of why Klingon -laH patterns with -lu'? QeS 'utlh
majQa'. vItIvqu'. ter'eS Sent from my Verizon, Samsung Galaxy smartphone -------- Original message --------From: Rhona Fenwick <qeslagh@hotmail.com> Date: 5/8/18 3:41 AM (GMT-06:00) To: tlhingan-hol@kli.org Subject: Re: [tlhIngan Hol] Moods and modality ghItlhpu' SuStel, jatlh:
Basically, how does one determine what moods a language actually has,
and can we apply this to Klingon?
My understanding is that "mood" as linguistically defined refers to specifically morphological means of signalling modality (that is, the speaker's subjective attitude towards the action of the verb as it is, should be, or may be) on finite verbs, since lexical expressions of modality are presumably limitless in any given language. Also, I think moods in a morphological system of modality would need to be mutually exclusive, otherwise you'd start multiplying moods together and end up with a much more complicated system. A challenge in distinguishing what constitutes mood and what doesn't is that the analysis varies from linguist to linguist; some linguists happily treat evidentiality, for instance, as a subset of modality, while others prefer to keep the two distinct (that's my perspective, since evidentiality is more about an objective statement of one's evidence for an assertion about what is, as opposed to one's subjective attitude towards what is, what should be, or what may be). You might like to read Ferdinand de Haan's discussion on evidentiality and modality (begins on p.47 of the PDF): https://pdfs.semanticscholar.org/f433/1055c9d78e29d7024f7d966fe22e5c749eac.p... Now, some tedious pondering on my part as to what this means for Klingon. Take all that follows with a good-sized grain of salt. With the idea of clearly separating evidentiality from modality, I'd probably treat Klingon as having four distinct morphological moods: one realis (indicative), two deontic (imperative, optative -jaj), and one epistemic (interrogative -'a'). The other Type 9s are all basically involved in making a verb non-finite, either as subordinate clauses (-DI', -bogh, -chugh, -pa', -vIS, -meH) or nouns (-ghach, -wI'), and because non-finite, modality doesn't really apply. Whether one would want to include the Type 2 suffixes in a count of moods is a more open question. I would tend to say no, because they seem to be able to happily co-occur with the non-indicative moods as well. In Turkish, -mAlI- forms (e.g. gelmeliyim "I must come") are often treated as a distinct necessitative mood, and the difference there is that a verb cannot be necessitative and any other mood at the same time: you can't have an imperative necessitative, for instance. Whereas in Klingon, as far as I know we can quite happily say something like mejnISjaj "may he have to leave" (even if the circumstances in which one would say something like that would be rather narrow). The Type 6 qualification suffixes fall pretty neatly into the category of evidentiality (and for me therefore not mood, though again, depends on one's analysis). For type 3, I think that's more aspect, though not the classic forms seen in the Type 7 aspect suffixes. -choH corresponds pretty neatly to what linguists call inchoative aspect, which is not uncommon in various languages. Much rarer is the idea of resumption that -qa' shows, but the Google overlords indicate there's a morphological resumptive aspect "again, starting again" in Kiliwa, a language of Baja California in Mexico. (Tangentially, Kiliwa was the dissertation topic of one Mauricio Mixco, who completed his doctorate at UC Berkeley, in the 1970s, and under the supervision of Mary Haas. Coincidence? I doubt it.) Type 8 (-neS) seems clearly just an honorific, with no other real semantic function (a couple of strange KGT examples aside). Type 4 (-moH), as a valency-changing device, would usually be referred to as voice (cp. other valency-changing devices like passive, middle, antipassive, etc.). Finally, Type 5. Here things get kind of interesting. -lu' isn't passive as such; we know that. But like the English passive, and indeed the Klingon causative, it's a valency-changing operation, therefore voice. -laH is a bit different because it doesn't actually alter valency in modern Klingon, so it's hard to call it voice sensu stricto. However, I've just discovered that in Japanese grammar what's usually called the "potential voice" originates from a form not unlike the passive, and has two case-marked forms, one of which does alter valency: Active: Tomoko ga mizu o nomimasu "Tomoko drinks water" Passive: Mizu ga (Tomoko ni) nomaremasu "water is drunk (by Tomoko)" Potential: Tomoko ga mizu o nomemasu "Tomoko is able to drink water" Potential: Mizu ga (Tomoko ni) nomemasu "water is drinkable (by Tomoko)" It may be worthy of noting that English deverbal adjectives in -able are generally passive in nature too: drinkable is not "able to drink", but "able to be drunk". Perhaps an insight (within the game, at least) into the historical origin of why Klingon -laH patterns with -lu'? QeS 'utlh
majQa'. vItIvqu'. ter'eS Sent from my Verizon, Samsung Galaxy smartphone -------- Original message --------From: Rhona Fenwick <qeslagh@hotmail.com> Date: 5/8/18 3:41 AM (GMT-06:00) To: tlhingan-hol@kli.org Subject: Re: [tlhIngan Hol] Moods and modality ghItlhpu' SuStel, jatlh:
Basically, how does one determine what moods a language actually has,
and can we apply this to Klingon?
My understanding is that "mood" as linguistically defined refers to specifically morphological means of signalling modality (that is, the speaker's subjective attitude towards the action of the verb as it is, should be, or may be) on finite verbs, since lexical expressions of modality are presumably limitless in any given language. Also, I think moods in a morphological system of modality would need to be mutually exclusive, otherwise you'd start multiplying moods together and end up with a much more complicated system. A challenge in distinguishing what constitutes mood and what doesn't is that the analysis varies from linguist to linguist; some linguists happily treat evidentiality, for instance, as a subset of modality, while others prefer to keep the two distinct (that's my perspective, since evidentiality is more about an objective statement of one's evidence for an assertion about what is, as opposed to one's subjective attitude towards what is, what should be, or what may be). You might like to read Ferdinand de Haan's discussion on evidentiality and modality (begins on p.47 of the PDF): https://pdfs.semanticscholar.org/f433/1055c9d78e29d7024f7d966fe22e5c749eac.p... Now, some tedious pondering on my part as to what this means for Klingon. Take all that follows with a good-sized grain of salt. With the idea of clearly separating evidentiality from modality, I'd probably treat Klingon as having four distinct morphological moods: one realis (indicative), two deontic (imperative, optative -jaj), and one epistemic (interrogative -'a'). The other Type 9s are all basically involved in making a verb non-finite, either as subordinate clauses (-DI', -bogh, -chugh, -pa', -vIS, -meH) or nouns (-ghach, -wI'), and because non-finite, modality doesn't really apply. Whether one would want to include the Type 2 suffixes in a count of moods is a more open question. I would tend to say no, because they seem to be able to happily co-occur with the non-indicative moods as well. In Turkish, -mAlI- forms (e.g. gelmeliyim "I must come") are often treated as a distinct necessitative mood, and the difference there is that a verb cannot be necessitative and any other mood at the same time: you can't have an imperative necessitative, for instance. Whereas in Klingon, as far as I know we can quite happily say something like mejnISjaj "may he have to leave" (even if the circumstances in which one would say something like that would be rather narrow). The Type 6 qualification suffixes fall pretty neatly into the category of evidentiality (and for me therefore not mood, though again, depends on one's analysis). For type 3, I think that's more aspect, though not the classic forms seen in the Type 7 aspect suffixes. -choH corresponds pretty neatly to what linguists call inchoative aspect, which is not uncommon in various languages. Much rarer is the idea of resumption that -qa' shows, but the Google overlords indicate there's a morphological resumptive aspect "again, starting again" in Kiliwa, a language of Baja California in Mexico. (Tangentially, Kiliwa was the dissertation topic of one Mauricio Mixco, who completed his doctorate at UC Berkeley, in the 1970s, and under the supervision of Mary Haas. Coincidence? I doubt it.) Type 8 (-neS) seems clearly just an honorific, with no other real semantic function (a couple of strange KGT examples aside). Type 4 (-moH), as a valency-changing device, would usually be referred to as voice (cp. other valency-changing devices like passive, middle, antipassive, etc.). Finally, Type 5. Here things get kind of interesting. -lu' isn't passive as such; we know that. But like the English passive, and indeed the Klingon causative, it's a valency-changing operation, therefore voice. -laH is a bit different because it doesn't actually alter valency in modern Klingon, so it's hard to call it voice sensu stricto. However, I've just discovered that in Japanese grammar what's usually called the "potential voice" originates from a form not unlike the passive, and has two case-marked forms, one of which does alter valency: Active: Tomoko ga mizu o nomimasu "Tomoko drinks water" Passive: Mizu ga (Tomoko ni) nomaremasu "water is drunk (by Tomoko)" Potential: Tomoko ga mizu o nomemasu "Tomoko is able to drink water" Potential: Mizu ga (Tomoko ni) nomemasu "water is drinkable (by Tomoko)" It may be worthy of noting that English deverbal adjectives in -able are generally passive in nature too: drinkable is not "able to drink", but "able to be drunk". Perhaps an insight (within the game, at least) into the historical origin of why Klingon -laH patterns with -lu'? QeS 'utlh
On 5/8/2018 4:41 AM, Rhona Fenwick wrote:
ghItlhpu' SuStel, jatlh:
Basically, how does one determine what moods a language actually has, and can we apply this to Klingon?
My understanding is that "mood" as linguistically defined refers to specifically morphological means of signalling modality
Indeed, I was going to bring up that distinction, which is why "modality" is in the subject line, but my topic was getting away from me, so I cut out the modality part.
(that is, the speaker's subjective attitude towards the action of the verb as it is, should be, or may be) on finite verbs, since lexical expressions of modality are presumably limitless in any given language. Also, I think moods in a morphological system of modality would need to be mutually exclusive, otherwise you'd start multiplying moods together and end up with a much more complicated system.
Does complexity disqualify combined moods?
With the idea of clearly separating evidentiality from modality, I'd probably treat Klingon as having four distinct morphological moods: one realis (indicative), two deontic (imperative, optative -*jaj*), and one epistemic (interrogative -*'a'*). The other Type 9s are all basically involved in making a verb non-finite, either as subordinate clauses (-*DI'*, -*bogh*, -*chugh*, -*pa'*, -*vIS*, -*meH*) or nouns (-*ghach*, -*wI'*), and because non-finite, modality doesn't really apply.
Noting that most of the syntactic suffixes make non-finite verbs is useful, and not something that had occurred to me. I see that none of your proposed moods can coexist, as you suggested.
Whether one would want to include the Type 2 suffixes in a count of moods is a more open question. I would tend to say no, because they seem to be able to happily co-occur with the non-indicative moods as well. In Turkish, -/mAlI/- forms (e.g. /gelmeliyim/ "I must come") are often treated as a distinct necessitative mood, and the difference there is that a verb cannot be necessitative and any other mood at the same time: you can't have an imperative necessitative, for instance. Whereas in Klingon, as far as I know we can quite happily say something like *mejnISjaj* "may he have to leave" (even if the circumstances in which one would say something like that would be rather narrow).
The Type 6 qualification suffixes fall pretty neatly into the category of evidentiality (and for me therefore not mood, though again, depends on one's analysis).
With my limited understanding, I tend to agree with the qualification suffixes being evidentiality and not mood. I'd like to see an argument against the type 2 suffixes beyond not being able to combine moods.
For type 3, I think that's more aspect, though not the classic forms seen in the Type 7 aspect suffixes. -*choH* corresponds pretty neatly to what linguists call inchoative aspect, which is not uncommon in various languages. Much rarer is the idea of resumption that -*qa'* shows, but the Google overlords indicate there's a morphological resumptive aspect "again, starting again" in Kiliwa, a language of Baja California in Mexico. (Tangentially, Kiliwa was the dissertation topic of one Mauricio Mixco, who completed his doctorate at UC Berkeley, in the 1970s, and under the supervision of Mary Haas. Coincidence? I doubt it.)
I unhesitatingly call the type 3 suffixes aspect.
Type 8 (-*neS*) seems clearly just an honorific, with no other real semantic function (a couple of strange KGT examples aside).
Type 4 (-*moH*), as a valency-changing device, would usually be referred to as voice (cp. other valency-changing devices like passive, middle, antipassive, etc.).
I've never heard of *-moH* being described as voice, but that's probably because I'm always so busy telling people *-lu'* isn't passive voice I hadn't considered other types of voice. You're absolutely right. According to Wikipedia, Mongolian, Turkish, Hebrew, and Japanese all have a causative voice. Very cool.
Finally, Type 5. Here things get kind of interesting. -*lu'* isn't passive as such; we know that. But like the English passive, and indeed the Klingon causative, it's a valency-changing operation, therefore voice.
I had to look this up to convince myself. There is a grammatical "fourth person," the indefinite person, used in languages like French and Welsh... and Klingon. But what I'm finding about it doesn't make clear the link between fourth person and voice. The fourth person is still the subject and agent, even if demoted in presence, and would seem to me to be active voice. In what way is valency changed in a *-lu'* verb?
-*laH* is a bit different because it doesn't actually alter valency in modern Klingon, so it's hard to call it voice /sensu stricto/. However, I've just discovered that in Japanese grammar what's usually called the "potential voice" originates from a form not unlike the passive, and has two case-marked forms, one of which /does/ alter valency:
Active: /Tomoko ga mizu o nomimasu/ "Tomoko drinks water" Passive: /Mizu ga /(/Tomoko ni/)/nomaremasu/ "water is drunk (by Tomoko)" Potential: /Tomoko ga mizu o nomemasu/ "Tomoko is able to drink water" Potential: /Mizu ga /(/Tomoko ni/)/nomemasu/ "water is drinkable (by Tomoko)"
It may be worthy of noting that English deverbal adjectives in -/able/ are generally passive in nature too: /drinkable/ is not "able to drink", but "able to _be_ drunk". Perhaps an insight (within the game, at least) into the historical origin of why Klingon -*laH* patterns with -*lu'*?
I don't see the connection here with Klingon, a completely unrelated language. You'd first have to convince me that *-lu'* is a separate voice, then maybe we could speculate on *-laH* being based in voice simply by virtue of its type 5 classification, even though it doesn't seem to change voice at all. Thank you very much for this! It's been extremely helpful to me and cleared up some confusion I've had about applying grammatical terminology to describe Klingon. -- SuStel http://trimboli.name
jIjatlhpu':
My understanding is that "mood" as linguistically defined refers to specifically morphological means of signalling modality
mujangpu' SuStel, jatlh:
Indeed, I was going to bring up that distinction, which is why "modality" is in the subject line,
Oh yes, that's entirely fair. In my original statement, I'd intended the emphasis on "morphological" rather than "modality". DopDaq qul yIchenmoH QobDI' ghu'. jatlhtaH:
Does complexity disqualify combined moods?
Again, unfortunately it really seems to depend on (a) the language and (b) the analysis. There's one analysis of Tundra Nenets that presents sixteen distinct moods (http://www.helsinki.fi/~tasalmin/sketch.html), but if I myself were analysing these, I'd be reducing them heavily on at least the basis of aspect as a second dimension (several of these alleged "moods" are presented baldly as morphologically complex aspectual distinctions within the same underlying mood). (poD vay') taH:
I'd like to see an argument against the type 2 suffixes beyond not being able to combine moods.
Me too, and while I was thinking about this I wasn't able to come up with any other decent argument for that. With that said, there are not many human languages that do similar things to Klingon in this respect, so I haven't been able to find any coherent pattern among linguists as to what to call such things. (Inuktitut seems to be the most studied language with similar forms: tikivuk "he arrives" has derived forms like tikigunnatuk "he is able to arrive", tikigumatuk "he wants to arrive", etc. But even in Inuktitut linguistics there doesn't seem to be a consistent terminology for these morphological units.) (poD vay') jIjatlhpu':
Type 4 (-moH), as a valency-changing device, would usually be referred to as voice
jang SuStel, jatlh:
I've never heard of -moH being described as voice, but that's probably because I'm always so busy telling people -lu' isn't passive voice I hadn't considered other types of voice. You're absolutely right. According to Wikipedia, Mongolian, Turkish, Hebrew, and Japanese all have a causative voice. Very cool.
Indeed. And I'm one of those who doesn't usually think of the causative as a distinct voice (even while Ubykh, my human-language focus, has a fully developed morphological causative); it's so often just called, unqualified, "the causative" and so I suppose I mentally file it under a different place, though I really shouldn't. taH:
In what way is valency changed in a -lu' verb?
Primarily in the sense that it has clear and striking effects on verbal agreement involving promotion of the patient to subject as well as demotion of the indefinite person to object agreement position. Hence why there isn't an effective contrast between wIqIplu' and *DIqIplu': the indefinite person - the notional subject - is always third-person singular object in terms of verbal agreement. It's not an increase or decrease in valency, but it's a valency exchange of sorts. The antipassive voice in Circassian behave similarly; because Circassian has trivalent agreement, an antipassive always conditions two verb agreement prefixes even when demotion occurs, but it involves argument exchange of almost exactly this sort, confirmed by the exchange of case-marking as well: active pχaʈʂ’em pχer jeχʷe "the carpenter (-m, j-) is filing the plank (-r, Ø-) to completion" vs. antipassive pχaʈʂ’er pχem jeχʷe "the carpenter (-r, Ø-) is filing away at the plank (-m, j-)". You're of course right that there's no effect on the argument structure from a syntactic point of view, and that (plus the retention of agreement for the indefinite person, albeit in object agreement position) is one major factor making it not a true passive, but I consider -lu' to be a form of voice - albeit an unusual one in terms of the clash between its syntax and its agreement pattern. Unfortunately, most languages with impersonal constructions of this kind don't have object agreement, so can't be of any terminological help; in the only one I've been able to find that does - Classical Nahuatl - the passive doesn't permit an unspecified agent to be resupplied by an adpositional phrase, and only the passive causes patient promotion to subject agreement (tamēchtlazohtlah "we (t-...-h) love you all (amēch-)", but antlazohtlaloh "you all (an-...-h) are loved"); no such promotion is permitted in the impersonal (tētlazohtlah "they (Ø-...-h) love someone unspecified (tē-)", but tētlazohtlalo "someone (Ø-...-Ø) loves someone unspecified (tē-)"). jIjatlhtaH:
-laH is a bit different because it doesn't actually alter valency in modern Klingon, so it's hard to call it voice sensu stricto.
jangqa' SuStel, jatlh:
I don't see the connection here with Klingon, a completely unrelated language.
I wasn't intending to suggest there's any direct connection at all. I tried to make it clear in my previous message that I'm suggesting a parallel, that's all, a parallel that only suggested itself through looking into why linguists of Japanese refer to this phenomenon as "voice". I'm just spitballing the possibility that if an unrelated, but similar, phenomenon occurred during the diachronic development of Klingon, that may be an explanation of why -laH and -lu' pattern together. (That's also why I specifically said I'm talking about within the game. I know it's not something that can be easily answered, but this isn't the first point of grammar that's made me think Okrand probably gave more than passing thought to giving Klingon an internal history when he built it.) taH:
You'd first have to convince me that -lu' is a separate voice, then maybe we could speculate on -laH being based in voice simply by virtue of its type 5 classification, even though it doesn't seem to change voice at all.
For sure, and from any synchronic point of view -laH is not voice in modern Klingon at all. Semantically it's really of a class with the Type 2s, but happens to appear in the Type 5s instead, whether due to some historical development or simply one of those kooky things that happens sometimes in languages. taH:
Thank you very much for this! It's been extremely helpful to me and cleared up
some confusion I've had about applying grammatical terminology to describe Klingon.
qay'be'qu'. And as I said, do take all the above with a healthy chunk of salt. QeS 'utlh
participants (3)
-
Rhona Fenwick -
SuStel -
terrence.donnelly