Tibetan Grammar - verbs - notes

From Rigpa Wiki
Revision as of 07:11, 30 March 2011 by Fone Bone (talk | contribs)
Jump to: navigation, search

WORK IN PROGRESS: the grammar articles are being edited for wiki publication. During editing, the content might be incomplete, out of sequence or even misleading. - (particular this sections is still in change - the introduction sections are all new - since the 'collection of points on Tibetan grammar' are now available outside of a class room context the background information to some of their points need to be written down, and this is still a work in progress)

Articles on Tibetan Grammar
1. Introduction
2. Formation of the Tibetan Syllable
3. Formation of the Tibetan Word
4. First case: ming tsam
5. agentive particle
6. Connective Particle
7. La don particles
8. La don particles—Notes
9. Originative case
10. Verbs
11. Verbs—Notes
12. Syntactic particles

by Stefan J. E.


This is written to explain the simplified way of classifying transitive and intransitive (1.1) and it points to the problems and inconsistencies coming with it (1.2). In 1.2 valency is introduced to map out Tibetan verbs, which might be not have been necessary as such, but seemed a nice way to do so and in its course introduces verbs that will be seen again in 1.4. "#'verbs and their cases', considerations and conclusion".

How the categories of 'transitive' and 'intransitive' are used here

In order to categorize Tibetan verbs according to their grammar the categories of 'transitive' and 'intransitive' will be used. The way it will be determined if a verb should be labeled 'transitive' or 'intransitive' will not entirely match the general rule for these categories.


  • Intransitive: Not passing over to an object; expressing an action or state that is limited to the agent or subject.
  • Transitive: Passing over to an object; expressing an action which is not limited to the agent or subject.
This section contains Tibetan script. Without proper Tibetan rendering support configured, you may see other symbols instead of Tibetan script.

The categorization will be in regard to the presence of an agent in the agentive case. In a number of cases this will lead to differences in regard to their English counterparts.

       For instance the English word "love" is transitive. There is 'somebody / thing' that is loved. In Tibetan "love" is an unintentional verb and has no agent marked with the agentive case (it is classified in as ཐ་མི་དད་པ་). Having these characteristics it will be categorized as an intransitive verb, in the category "verbs of emotion / attitude verbs" and its grammar described as:
Patient (subject): ming tsam, and qualifier—that which the attitude is towards: la don.

shepherd sheep   kind, loving
The shepherd is loving to the sheep.

In most cases this way of dealing with Tibetan verbs leads to a straight forward way of categorizing them. Yet it does lead to problems with some 'transitive verbs with la don' (see below) and can obscure the fact that divalent intransitive verbs are simply the unintentional counterpart of intentional transitive (divalent) 'verbs with la don' (see below).

Classification as Patient, subject-object: advantages and problems

This section is written in regard to the simplified way of classifying transitive and intransitive, (see above). It points to the problems and inconsistencies coming with it.


  • Patient here is used as a convenient term for the
  1. subject of an intransitive verb and the
  2. object of a transitive verb.

These two are mostly in the ming tsam case—marked by no particle—'just the word'.[1] The term Patient is stretched beyonds its definition from thematic relations; e.g. it will also include theme—undergoes the action but does not change its state, and experiencer—the entity that receives sensory or emotional input. Patient is also used with static verbs.[2]

In general the patient is that which experiences the action. In many cases[3] it is equal to the object of a transitive verb. The difference between it and an object is that patient is based explicitly on its relationship to the verb, whereas the object is defined primarily through its relationship to the subject.

In Tibetan where the type of verb governs the usage of the respective particles for their agent, patient and particular qualifiers it can be fitting to use these verb dependent categories (of patient and agent) in order to describe the grammar of verbs.[4]

Moreover it is much easier to explain Tibetan using a single term that covers the subject of an intransitive verb and the object a transitive verb. In Tibetan the patient is in 90+% of all cases in ming tsam, which makes the use of "patient" an advantage for beginners. It is easy to keep in mind that one needs to look for 'something' in ming tsam in order to find the patient of the clause / sentence. (Whereas looking for the subject of a transitive verb could be quite disheartening, given that it is so often omitted.)

In the most part it is straight forward to classify the grammar of verbs using the cases in which their patient, qualifier and agent, if present, are in. It is also easy to describe verb-verb relations in terms of a verb with either a patient (complement) or a qualifier.

However some verbs are problematic when using 'patient'. In order to see where these problems come from there will be an overview of Tibetan verbs with an attempt to use valency as a way of ordering them.

Tibetan verbs in a valency matrix


The term valency or valence[5] refers to the property of a word 'to bind' other words to it, 'to demand' complements. The study of valency structures can be quite detailed.[6]

The concern here is the obligatory complements. Obligatory complements are complements which have to be expressed in a grammatical sentence to enable the use of the predicator (verb), the verb requires all of the arguments (complements) in a well-formed sentence. However verbs sometimes undergo valency reduction or expansion.[7]

Types of valency

  1. Monovalent verb takes one argument, e.g. "He sleeps."
  2. Divalent verb takes two, e.g. "He hit the king."
  3. Trivalent verb takes three, e.g. "He gave her a ring."
  • Zero valency: When a complement status is not attributed to "it" (even though that it is certainly to be regarded as a property of the governing verb that it takes "it" as its subject) then one needs to add zero valency, the avalent verb that take no arguments, e.g. "It rains.", with the explanation that "it" is only a dummy subject and a syntactic placeholder—"it" has no true meaning. No other subject can replace it.

Valency and Tibetan

Valency comes from the study of languages that generally don't have the ability to omit the same amount of components of a sentence as Tibetan does. In Tibetan a sentence does not become ungrammatical or poorly formed by omitting parts that are to be understood from context, even if it is the subject or object of the sentence. It could even be bad style to state them.

For that reason the way valency will be used here is to look at the number of obligatory complements of a sentence without omissions. When counting obligatory complements there might be some debate with questions like "What can't be left out?" and "What needs to be always assumed?". For instance with the verb "to look" is it assumed that there is always something which is looked at? If it is, with verbs of living is it assumed that there is always a place where one stays?

The valency model is used here as merely an aid to illustrate the main differences between Tibetan verbs, with the 'divalent verbs with la don' as the main topic. In this context the verbs of perception are treated as divalent whereas verbs of living and motion as monovalent.

Note: The category of zero valency (see above) which is used in the great work "Lhasa Verbs"[8] will not be used. Compound verbs like ཆར་པ་འབབ་ "to rain" will be treated as the monovalent verb འབབ་ "to fall" with the noun ཆར་པ་ "rain".

Tibetan verb valency-particle-volition matrix

This section is in particular about transitive verbs with their patient / object marked by la don. For that reason other divalent verbs like verbs of separation that have a qualifier-what one is separated from-are excluded.

The examples:

sun   arose
The sun arose.
they went
They went.
shepherd sheep   kind, loving
The shepherd is loving to the sheep.
nectar  like Dharma a/one  I   found
I have found this nectar like Dharma.

Buddha      Dharma taught
The Buddha taught the Dharma.
he     book(s)  looked
He looked at books.
I          other  benefit  will (aux)
I will benefit others.
doctor  the ill  medicine give
The doctor gives medicine
to the ill.

སྙིང་ནས་གྲོལ་བ་དོན་དུ་གཉེར་བའི་གང་ཟག་གིས། བདག་མེད་པའི་ལྟ་བ་རྣམ་པར་དག་པ་ཁོང་དུ་ཆུད་པའི་ཐབས་ལ་འབད་དགོས།
heart  liberation    seek             persons        selflessness    view completely pure  understand  means  effort  need
Persons who from the depths of their hearts seek liberation must work at the means of understanding the correct view of selflessness.

valency type of verb agentive case ming tsam lad don verb example
monovalent unintentional,
patient / subject
intransitive verbs
ཤར་ (ཐ་མི་དད་པ་)
patient / subject
verbs of motion
སོང་ (ཐ་མི་དད་པ་)
divalent unintentional,
patient / subject
attitude verbs
བྱམས་ (ཐ་མི་དད་པ་)
agent / subject
patient / object
"fruitional" verbs
རྙེད་ (ཐ་མི་དད་པ་)
agent / subject
patient / object
transitive verbs
བསྟན་ (ཐ་དད་པ་)
agent / subject

agent / subject
"patient / object"

"patient / object"
transitive verbs
བལྟས་ (ཐ་དད་པ་)

verbs of benefit
ཕན་ (ཐ་མི་དད་པ་)
agent / subject
verbs expressing
"to make effort"
འབད་ (ཐ་དད་པ་)
trivalent intentional,
agent / subject
patient / object
recipient / indirect object
ditransitive verbs
སྟེར (ཐ་དད་པ་)

* Because this section puts the simplified approach in more grammatical context, some of the labels "transitive" and "patient / object" are in inverted commas, indicating that there is more to those.

When looking at which particle marks what, the agentive case always marks an agent, and ming tsam always a patient, so why do the la don seem to be 'multitasking', marking qualifier for unintentional and patient for intentional verbs? Or, is there even a real difference between what they are marking for unintentional and intentional divalent verbs?

Divalent verbs with la don

La don do have a wide range of functions, but they all fall into the category of marking some kind of qualifier. This comes from their origin of being words of location and direction.[9] Which would make the complement of the transitive divalent verbs the 'direction' the action is directed towards and not the patient.

For example, Peter Schwieger treats these verbs as intransitive, pointing out that even though they are ཐ་དད་པ་ classified they have their object marked with the la don and are not transitive.[10] The examples are འཛེག་, "to climb", བརྩོན་, "to strive" and གནོད་, "to harm". The {{gtib|ཐ་དད་པ་{ verbs ལྟ་བ་, "to look", གནོད་པ་, "to harm" which both use la don are placed with 'intentional (controllable) intransitive verbs'.[11]

The examples:

then      Ānanda       direction    four   looked
Then Ānanda looked into the four directions.
that  who harm  who   harm
Who harms that [one]? Who is harmed?

When looking at those unintentional and intentional divalent verbs one can say that they are the unintentional and intentional counterpart of each other.[12] With that view it follows logically that the agent of the intentional verbs would be the one doing and experiencing the action just as the patient of the unintentional verbs does. The agent would substitute the patient, or be the patient.

Being the patient or agent should be mutually exclusive and is only possible here due to the way these terms are used as terms of convenience. The patient being the one undergoing the action, and the agent the one marked with the agentive case. (Ts. Takeuchi and Y. Takahashi's conclusion about agent and experiencer of transitive verbs is:"... subjects in transitive sentences have the ergative marker, be they agents or experiencer,..".[13] - Their conclusion is that the experiencer can be marked with the ergative / agentive case.)

When following the convention here that the agent in the agentive case, then one option is to take the agent as the substitute for the patient. That is done in the case of "verbs expressing 'to make effort, to engage'". It is not used with any of the other intentional divalent verbs. The next sections explains why.

Intentional divalent verbs with la don and the 'labeled' patient

Keeping with the way of labeling verbs described above, all verbs that have an agent marked with the agentive case are categorized as transitive. As a result the intentional divalent verb are labeled as transitive too. In most cases this label is appropriate because the majority of these verbs have a complement that one would usually consider to be an object rather than a qualifier.

A generic example:

he          dog   see
He sees the dog.
he           dog  look
He looks at the dog.
divalent, unintentional,
ཐ་མི་དད་པ་ classified
divalent, intentional,
ཐ་དད་པ་ classified

It takes some explaining in order to show how the dog that is seen is more involved in the action than the dog that is looked at.[14]The 'looked at dog' is not in ming tsam and there is a grammatical difference. The 'seen dog' is an object of a transitive verb and 'looked at dog' is the direction of focus for an intransitve verb (see below). Yet in a pragamaitc context the diffetence becomes much less apparent. With "verbs of harm" like གནོད་པ་, "to harm" that the complement 'that what is harmed' is directly involved in the agent's "performance" of the verb. (See blow.)

The reason why the degree of involvement of the complement with the action of the verbs differs, comes form the fact that intentional divalent verbs with la don are in themselves not one type of verb. They range from "verbs of benefit and harm" with their complement having all the characteristics of a patient, "intentional verbs of perception" and "verbs expressing mental activity" with an adverbial complement that could be viewed as an object or patient, to the other side of "verbs expressing 'to make effort, to engage'" which come with an adverbial complement that shows the direction of the action of the verb without any involvement in it. This range is also illustrated by their English counterparts. The "verbs of benefit and harm" are transitive,[15] e.g. "She benefits the school.". The "intentional verbs of perception" and "verbs expressing mental activity" are not as clearly cut, they can be either transitive or intransitive or be used in both ways, e.g., v.t. "She thinks virtuous thoughts", v.i. "He thinks about leaving" and the "verbs expressing 'to make effort, to engage'" are intransitive, e.g. "She strives for success.".

Here, when the complement marked by la don has qualities of an object it will labeled as 'patient' (which is marked by la don). In the example with "look" that what is looked at "dog" will be labeled as the patient.

This does not work for "verbs expressing 'to make effort, to engage'". There simply comes a point when even the very enduring 'patient' can not be stretch any further. That 'what the effort is towards', the adverbial complement that shows the direction of the action, is a qualifier marked by a la don and not a patient. In order to keep with the way of categorizing—that the agents with the agentive case is the sign of a transitive verb—the "verbs expressing 'to make effort, to engage'" are labeled "intransitive verbs with transitive grammar".

Conclusion: advantages vs. problems

This simpliefied transitive - intransitive categorisation by way of the agent is for the classroom. It leads to having all agents (obviously) in the agentive case and 90+% of all patients in ming tsam, and for some transitive verbs the patient marked by la don. The main point is, that it has proven to be a very comprehensive approach for students encountering classical Tibetan as beginners. (See also: Introduction, origin and aims of this 'collection of different points on Tibetan grammar').

It does label some verbs that are actual intransitive as transitive and does lead to the above explained inconsistencies of some patients being marked by the la don and the "verbs expressing 'to make effort, to engage'" needing an extra category.

Another reason to label intentional divalent verbs with la don as transitive due to the occurrence of an agent is because the effectiveness of using the agent basis for the categorisation, while at the same time other ways to categorize transitive and intransitive verbs come with their own problems:

  • Taking the Tibetan categorization of ཐ་དད་པ་ and ཐ་མི་དད་པ་ as a bases leads to trouble with ཐ་མི་དད་པ་ classified verbs with transitive grammar such as unintentional verbs of perception (intentional divalent verbs would still be transitive because they are ཐ་དད་པ་ verbs).
  • Looking at the presence or absence of an object leads to the question why with "to love" བྱམས་པ་ "that what is loved" is not the object but a qualifier and with "to harm" གནོད་པ་ "that what is harmed" is not the object (it corresponds very well to the definition of patient-the participant that suffers, endures the action).
  • Using the occurrence of a patient in ming tsam allows for clear categorization but it leads to the fact that verbs like "to harm" གནོད་པ་ would be intransitive verbs with an agent in the agentive case and a qualifier that looks like a patient /object. (Using the patient in ming tsam would probably my approach of choice, (including an exception for 'verbs of benefit and harm'), but at this point I still think that the approach by agentive case agent is the easiest for learners. This is clearly a choice that could be debated.)

Because of verbs like "to harm" གནོད་པ་ where it looks very appropriate to treat them as having a patient marked by a la don, the reason that this way of explaining verbs is very comprehensible for students and the fact that intentional divalent verbs are ཐ་དད་པ་ verbs in Tibetan this way of categorizing of verbs is chosen—placing 'intentional divalent verbs with la don' with transitive verbs. Provided that it is clear what this classification is based on, why there are occurrences like "verbs expressing 'to make effort, to engage'" that need their own category, it will hopefully be considered reasonable.

Volition: transitive-intransitive

Volition in general refers to a distinction that is made in a verb's conjugations[16] or case assignment[17]to express whether the subject intended the action or not, whether it was done voluntarily or accidentally (involuntarily).

In Tibetan volition does neither entirely rule the usage of the agentive case nor he distinction between ཐ་མི་དད་པ་ and ཐ་དད་པ་:

  • Unintentional intransitive verbs are ཐ་མི་དད་པ་ and have no agent in the agentive case.
  • Intentional intransitive verbs are ཐ་མི་དད་པ་ and have no agent in the agentive case, i.e., verbs of motion and living.
  • Unintentional transitive verbs are ཐ་མི་དད་པ་ and have an agent in the agentive case, e.g., unintentional verbs of perception.
  • Intentional transitive verbs are ཐ་དད་པ་ and have an agent in the agentive case.
volition intransitive transitive
unintentional E.g. unintentional intransitive verbs, e.g. འཆར་བ་, to appear
unintentional verbs of feeling, e.g. བཀྲེས་པ་, to be hungry
E.g. unintentional verbs of perception, e.g. མཐོང་བ་, to see
verbs of "understanding", e.g. ཧ་གོ་བ་, to understand
"passive / fruitional" verbs, e.g., འཐོབ་པ་ to attain, to obtain
ming tsam ཐ་མི་དད་པ་
patient / subject
intransitive verb

The flower blossoms.

agentive case ming tsam ཐ་མི་དད་པ་
agent / subject
patient / object
"fruitional" verb

Ananda attained [the state of an] arhat.

intentional E.g. verbs of motion, e.g. འགྲོ་བ་, to go
verbs of living, e.g. སྡོད་པ་, to stay
E.g. intentional transitive verbs, e.g. སྟོན་པ་, to teach, 
འཐུང་བ་, to drink, བཟོ་བ་, to make, do, produce, manufacture
ming tsam lad don ཐ་མི་དད་པ་
patient / subject
verb of motion

He goes to Lhasa.

agentive case ming tsam ཐ་དད་པ་
agent / subject
patient / object
transitive verb

He drinks tea.

'verbs and their cases', considerations and conclusion


  1. S. V. Beyer: The Classical Tibetan Language, p.259-260: "Intransitive verbs occur with a patient; transitive verbs occur with both a patient and an agency. [...] Tibetan—syntactically identify the intransitive and transitive patients. In Tibetan they both given the patient role particle.
  2. In S. V. Beyer's approach, ibid., p.263: "The patient of an event is the participant that suffers, endures, or undergoes the particular state, process, or action; the patient is the one the event happens to"
  3. It is for instance not the case in English passive constructions. For example, in the active voice phrase "The snow leopard bites the dog", the dog is both the patient and the direct object. By contrast, in the passive voice phrase "The dog is bitten by the snow leopard", the dog is still the patient, but now stands as the phrase's subject; while the snow leopard is only the agent.
  4. This is far less useful, if at all, for spoken Tibetan where the subject is the ruling factor for the auxiliary verbs and with the occurrence of a fluid-S Split ergative in regard to the degree of volition.
  5. The linguistic meaning of valence derives from the definition of valency in chemistry, where it is is a measure of the number of bonds formed by an atom of a given element.
  6. There can be the quantitative and qualitative syntactic and semantic valency, and categories of obligatory complements, optional complements, contextually optional complements, and adjuncts.
  7. E.g.: Divalent "He is drinking a coffee.", may be reduced to monovalency in "He is drinking."
  8. Lhasa Verbs, A Practical Introduction, by Tibetan contributors: Pema Gyatso, Dawa, Dekyi; created and produced by: Geoff Bailey and Christopher E. Walter
  9. Walter Simon: Certain Tibetan Suffixes and Their Combinations, Harvard Journal of Asiatic Studies, Vol. 5, No. 3/4. (Jan., 1941), p.378 ff. "3. la: There is an Old Tibetan form laḥ for la, as we had naḥ for na, and there is the verb lan̂ "to rise," a secondary form of ldan̂. Does then la-lan̂ form a pair to match na-nan̂ and ste-sten̂? I believe it does, [...] To discuss finally the meaning for la as a suffix, there can be no doubt that it has acquired a very generalised meaning, but I believe that the meaning "above, upon, on top" can still be felt in many cases..." See p.380 ff. for "4. du (tu, r(u), su)"
  10. Peter Schwieger: Handbuch zur Grammatik der klassischen tibetischen Schriftsprache p.75, "Häufig wird die Unterterscheidung von ཐ་དད་པའི་བྱ་ཚིག་ und ཐ་མི་དད་པའི་བྱ་ཚིག་ pauschal mit der Unterscheidung von transitiven und intransitiven Vernen gleichgesetzt. Diese Gleichsetzung trifft zwar in vielen Fällen zu, ist pauschal aber unzulässig. Es finden sich nähmlich eine Reihe von Verben, die zwar als ཐ་དད་པ་ klassifiziert werden, i.d.R. aber die Suffigierung ihres Objectes mit la oder tu bzw. seinen Allomorphen fordern...."
    "The differentiation between ཐ་དད་པའི་བྱ་ཚིག་ and ཐ་མི་དད་པའི་བྱ་ཚིག་ are often treated to be the same as the differentiation between transitive und intransitive verbs. While this equivalence applies in many cases it is not possible to use it as a general rule. Because there are a number of verbs that even though classified as ཐ་དད་པ་ require the suffixation of their object with la or tu and its respective allomorphs."
  11. Ibid., p.77, "b) kontrollierbare intransitive Verben"
  12. This is for the purpose of looking at the function of the la don and just a very rough way to look at these verbs. There is an ongoing interest in research in and debate about the origin and meaning of the Tibetan agentive case and its usage. The unintentional and intentional divalent verbs with la don may have quite different origins and the agentive case might not even marking a volitional action. See below.
  13. Tsuguhito Takeuchi and Yoshiharu, Osaka, Japan 1995, Split Ergative Patterns in Transitive and Intransitive Sentences in Tibetan: a Reconsideration: "Tibetan shows, both in its historical forms and in the Modern Central dialect, a relatively plain ergative marking pattern, where, except for 'marked' usages, subjects in transitive sentences have the ergative marker, be they agents or experiencer, and those in intransitive sentences have no marker."
  14. An object in grammar denotes somebody or something involved in the subject's "performance" of the verb. Basically, it is what the verb is being done to. There is of course a difference between "to look" and "to see". The difference that occurs in Tibetan is also seen in English—"to see" is transitive with its object in the objective / accusative case (still visible with "him" and "her"—which is same as the dative case in English) and "to look" is intransitive with an adverbial prepositional phrase as complement starting with "at", "into", etc.
  15. Transitive when meaning "to be helpful or useful to". They are intransitive in English when meaning to derive benefit, e.g. "He benefits from her good example."
  16. Conjugation is the creation of derived forms of a verb by inflection (regular alteration—change of sound/spelling—according to rules of grammar).
  17. Like assigning the agentive case to the subject of an intentional verb.