Tenses, Aspects, and Categories of Evidentiality and Egocentricity in Spoken Lhasa Tibetan

Yue Ming  Abstract: This paper mainly discusses verbal aspects and their syntactic markers in Lhasa Tibetan. Since aspects in Tibetan always share markers with other important categories called evidentiality and egocentricity, the former of which assesses the evidence of speaker's statement and the latter denotes behavioral characteristics of a speaker's ego-centricity, this paper simultaneously describes all three types: aspect, evidentiality and egocentricity. There are nine types of verb aspects in Lhasa Tibetan, namely: Prospective, Imminent, Premediated, Realis, Durative, Resultative, Merely-past, Perfect, and Contextual aspects. There are four types of evidentialities, namely, Self-knowledge, Knowing-by-sight, Knowing-by-newly-found-information and Knowing-by-inference evidentialities. Categories of egocentricity demonstrate whether an action is towards the Self or the Other, whether the speaker gains or loses, and so on. Additionally ， this essay also discusses the interactive relationship between syntactical verb aspects and personal pronouns, intentionality, modality, as well


Introduction
This article discusses syntactic phenomena around categories of verb aspects, evidentiality and egocentricity in the Lhasa dialect of Tibetan, focusing on verb aspects as its main subject. To make things easier, conventional names for most types of the verb aspects discussed will be used in the essay. As for the few innovated terms, or terms used Tibetan verbs and verb phrases involve a wide range of syntactic categories, including Person, volitionality, modality, spatial-temporal relation, reference, narrative viewpoint (ego/non-ego), speech evidence (evidentiality), and so on. The formation of these concepts is closely related to Tibetan people"s life experiences and perception of the physical world. Among others, the narrative viewpoint reflected in egocentricity is a principal concept, which is linguistically represented by experiential classification of grammatical functional markers, as well as predicate verbs. Owing to the diversity of the experiential world, the classification is indeed a complicated interaction of categories. For instance, subjective volitionality can be dependent or independent; egocentric perception can be either an extrovert or introvert referential behavior; first-hand experience can be determinant or non-determinant; events can be of various progressive states under self-observation：just started or already finished, durative or realized; or of various evidential properties based on the evidences produced during the speech. All these notions are represented more or less through various syntactic forms of verbs or verb phrases (though some forms carry certain semantic meanings), and some have even been grammaticalized. Verbal aspects, evidentiality and their syntactic formal markers discussed in this essay reflect the linguistic representation of the Tibetan people"s view of the world.
This essay holds that there are basically nine types of "verbal aspect" in Tibetan， namely：the Prospective aspect, Imminent aspect, Premediated aspect, Realis, the Durative aspect, Resultative aspect, Merely-past aspect, Perfect aspect, and the Contextual aspect. All these syntactic aspects have their own formal markers. On the other hand，Tibetan categories of aspects always share their formal markers with evidentiality and egocentricity. Hence this paper will also describe evidentiality types which evaluate the speaker's evidence and egocentricity types which reflect the speaker's egocentric behavior. ②

Prospective aspect
The Prospective aspect in Lhasa dialect may also be called the Premeditated aspect. It indicates a premeditated action (or event) closing-up to the scheduled time of its happening. This is of course quite subjective. The Pprospective aspect takes "gi yin" and "gi red" as its ② This essay uses the following signs: C=connective;m.w=mood word;tap=target particle;cop=complement particle;cau=causative particle;pap=pause particle;hop=honorific particle;stc=state complement;agn=agentive;pos=possessive;loc=locative;dat=dative;obj=objective;ala=allative;gen=genitiv e;ins=instrumental;abl=ablative;pro=prospective aspect;imm=imminent aspect;prm=premeditated aspect;rea=realis;dur=durative aspect;mer=merely past aspect;rst=resultative;con=contextual;pef=perfect aspect;eg=self-knowing (ego);sf=knowing-by-sight;ni=knowing-by-newly-found-information;jd=knowing by inferential;nom=nominization. Please refer to the Appendices at the end this paper. For more details, see Jiang Di,2004. formal markers，the former suggesting the narrator"s desire to show his/her confidence and grounds, and the latter suggesting the narrator"s estimation or speculation about other people"s behavior according to his/her experiences. From this pragmatic perspective, there is some evidentiality differences between "gi yin" and "gi red", with the former being a formal marker of self-conscious evidentiality in the prospective aspect (pro.eg) and the latter a formal marker of extrapolated evidentiality (pro.jd) (also see Appendix II) .
I C hometown-GEN folks-AG you-OBJ welcome-PRO.EG Folks and I welcome you.
Tomorrow schoolmates ball play-PRO.JD Schoolmates are to play a ball game tomorrow.
According to the Tibetan syntactic rules, the subject of a sentence in prospective aspect usually has an agentive tag, but in Lhasa dialect these markers are not compulsory. Agentive tags are added only when special emphasis is needed or when there is the possibility of misunderstanding the sentence.
You-GEN kindness that I-AG forget-NEG-PRO I will never forget your kindness.
The Prospective aspect in Tibetan entails some presupposition to the state of the occurrence of an event. If such occurrence becomes a convention or inevitable consequence，then new syntactic signification can be derived, including a relationship of cause-and-effect speculation, and lasting status of the occurrence of the event.
Sun east-ABL rise-PROSE west-ABL set-PRO The Sun rises in the East and sets in the West.
While "gi yin" is usually used after volitional verbs, "gi red" can be used after both volitional and non-volitional verbs due to its representation of habitual actions or inevitable physical phenomena.
pride C fall behind-PRO Take humans as an example，modesty helps one go forward, whereas conceit makes one lag behind.
Medicine a little take C pain the relieve can-PRO Take some pills, and the pain shall be relieved.
The prospective aspect marker "gi red" can also be used after adjectival predicates for comparison.
As to the syntactic function of "verb+gi yin /gi red" pattern in Lhasa Tibetan, C. A. Bell (1905:47),Jin Peng (1983:45), Zhou Jiwen (2003, and P. Denwood (1999:155) all regarded it as a marker of tense, while Qu Aitang and Tan Kerang (1983) regarded it as a marker of the prospective aspect. We agree with Qu and Tan.
In addition, "gi yin" has a function similar to that of "gi red" in marking the status of prospective events, yet the meaning of habitual actions and inevitability that "gi red" has evolved into has no match in "gi yin". This is because the meaning of habit or inevitability is an opposite notion to the speaker's volitionality. Consequently, "verb＋gi red" as a form of inevitability may also be excluded from the prospective aspect to form a category of customary aspect, which will be discussed in another essay.
The verbal ending "mkhan yin／mkhan red" can also be used in the prospective aspect in Lhasa dialect. "yin／red" was originally a lexical suffix attached to verbs to indicate the agent, while "yin/red" was originally a copular verb. The marker may have been grammaticalized from this type of combined copular sentences.

You
where go-PRO-M.W Where are you going.
You beer buy-TAP go-PRO-M.W Are you going to buy some beer.

Imminent aspect
There are three formal markers， "grabs yod"， "grabs dug"， and "grabs yoo red" for the Imminent aspect. "grabs" in Lhasa dialect is pronunced as [], or "groo" or "groo" in writing. The Imminent aspect means an action is about to happen. It does not usually imply premediation as the prospective aspect does.
sun sunset-IMM NOM landlord-DAT give The carpenter made a horn cap, and gave it to the landlord when the sun was about to set.
Jin Peng (1983:46) regarded "grabs yod","grabs dug",and "grabs yoo red" as markers of an Imminent Tense, with the first two markers differentiating from the last marker in mood (special declarative mood vs. general declarative mood) , and the difference between the first two lies in the narrator"s evidentiality, that is, a difference of whether the narrator has knowing-by-sight affirmation or can only make a subjective judgment. Yet future events are anyhow different from those events happened, and therefore always bear a color of being speculated, no matter how strong the evidentiality is. Sentence (11) with "grabs vdug" is a speculation based on some experiences, while sentence (13) with "grabs yod" has much certainty (that schools usually start at 8 o'clock in the morning). Generally speaking, there are many deep-rooted factors that influence the selection of aspectual markers in Tibetan speech, including the speaker's volitionality and degree of concern, the time and place of the event occurred, the relation between the speaker and the event, and so on.
The difference between volitional and non-volitional verbs in sentences of imminent aspect also suggests certain objectivity of the event occurrence. In regard to volitional verbs， the speaker represented by self-reference may be volitional, but may also entail other factors such as responsibility, obligation, convention, etc. For instance， (15) da-dus-chod slebs-grabs-dug. nga kha-la bzo-grabs-yod.

Time
arrive-IMM I meal cook-IMM It's time for me to cook the meal.

Cat this fish catch-IMM
The cat is about to catch some fish.
His-GEN name I-AG forget-IMM I have almost forgot his name.
The multiple markers of the Imminent aspect entail many meanings, as the occurrence and existence of multiple forms must be driven by multiple functions. At least three hierarchies can be distinguished from the interrelationship between those functions and markers. In Table 1, the deepest concept is narrative viewpoint, which is related to human cognitive experiences in early enclosed societies and is recorded in language. The narrative concept has a cultural heritage principally embodied in "egoness" and "nonegoness". "Egoness", is an "ego-centric point of view" which takes the real world as a context around the speaker; whereas "non-egoness" indicates a "non-egocentric" or "other-centric" viewpoint (see also the discussion on "Perfect apsect"). The Evidentiality layer, however, reflects how the speaker observes the world and how he categorizes his evaluation of events, for Tibetan people believe that speech evidentiality must be overtly expressed in the discourse. Such parameters entail content semantic and logical conditionality，among which Person and verb volitionality not only interact with other syntactic elements in the discourse, but also relate to discursive viewpoint and experiential knowledge. Furthermore, self-reference means using a first-person pronoun, while other-reference includes second-person and third person pronouns. The benefit of using such a classification of Person was discussed in detail by Sun Tianxin (1995) from a pragmatic conversational principle and evidentiality perspective. This paper also adopts this classification.
Of course there will be some exceptional cases, such as those pointed out by Zhou Jiwen (2003:106): "grabs yoo red" can be used to mean "the elapsed future", for instance, "chu tshod drug par (six o'clock) slebs (arrive) shag. khong (he) slob grwai nang la (at school) sdod (stay) dus (time). da (now) lang (get up) grabs yoo red." (It's six o'clock. He would have got up when he was at the boarding school.). The notion of time is included here. And last but not least, the time of the event occurrence will absolutely be influenced by the person"s experiential knowledge. Because the evidentiality of future events is mainly based on previous experiences, and life experiences differ from person to person, case to case, there will be some free variations in the evidentiality layer, which will consequently lead to some chaos in the selection of narrative viewpoint, and finally to the collocation and marking of the parameters. These phenomena may also interact with the development of the language. Apart from those imminent aspect markers discussed above，there is another expression in Lhasa Tibetan to express the occurrence of events, that is, "Verb＋grabs byed kyis". The verb in this pattern can be reduplicated, as in: (18) de-dus shar-phyogs-nas nyi-ma dmar-chem-chem shar-grabs-shar-grabs-byed-kyis.
Then orient-ABL sun red rise(-IMM)-DUR The red sun was about to rise from the East.
However, it may not be proper to regard this expression as a syntactic aspectual marker in Tibetan.

Premediated aspect
This aspect is unique in that it has a dual meaning: the event is about to happen but has not yet happened. In other words,it entails the inevitability of the occurrence of the event, while denying the happening of the event. Consequently, this dual meaning is represented as syntactically-semantically known. The Premeditated aspect suggests some objectivity of the occurrence of an event, such as the occurrence of objective needs, customs, regulations, etc.
Verbs in the Premediated aspect are added with "rgyu yin" or "rgyu red", depending on whether the speaker makes his own decision or he speculates on other's behavior. As analyzed above, when the speaker volitionally decides his own actions, pragmatically it's an egocentric behavior, and this egocentric viewpoint takes the marker "rgyu yin"，rather than the non-egocentric "rgyu red", the non-volitional self-reference pattern included. For instance, sentence (20) implies the meaning that "they asked us to help the chef ". (20) nga-tsho-s las-ka ga-re byed-rgyu-red.
today we chef-GEN help-PRM.
What are we going to do？ We are to help the chef.
Sentences in premediated aspect cannot be in the negative form, for the aspect indicates the event is yet to happen. (22) bgra-shis kha-lag za-rgyu-red-pas.
doctor good in this way-NEG C I die-M.W Had I not seen such a good doctor, I would certainly have died.
Scholars are divided on the categorization of this structure. Zhou Jiwen (2003:108) considered it to be a marker of the future tense, while P. Denwood (1999:233) regarded "rgyu" as a marker of nominalization. We agree with Denwood,and exclude "rgyu rang red" from the markers for the Premediated aspect.

Durative aspect
The Durative aspect in Lhasa Tibetan is used to refer to ongoing actions or actions that are repeated during a certain period of time. Durative events, durative states, and customary conventions are also included.
"verb＋gi yod／gi dug／gi yoo red" are formal markers for the Durative aspect. "gi yod" is usually used to indicate the self-reference of the speaker"s volition, or the closeness or relatedness to the speaker"s volition, while "gi dug" and "gi yoo red" are used in cases of other-reference. Most action verbs clearly indicate an ongoing state of an event, suggesting that the action is still in a durative progress.
girl snag-sa-AG crowd-GEN middle-ABL Itad-OBJ see-DUR Snag-sa is in the crowd watching Itad.
Suggesting the ongoingness of an event, the Durative aspect does not entail any ending point. Hence, many general statements, or philosophical/scientific statements, and proverbs such as "the Earth moves round the Sun" in English，are all of this feature. Some static verbs and weak action verbs have a similar feature, with but a limited time span, like in "She knows several languages". Therefore, the Durative aspect indicates not only the progress, but also the status of an event, a custom, and certain ability in a given period of time. It indicates the durativity of an event, but not the temporal past, present or future.

Summer
hot-DUR-M.W Is the summer hot?
Tshwar-dam region-LOC salt exuberance have so not take C outflow-DUR Salt is abundant in the Tshwar-dam region，it comes out itself without any labour effort.
The convention of an event or action repeating itself regularly (every day/week/year/etc.) is a state of the event in that given period. A fixed lifestyle, or an equivalent exchange, is a state as well.
Lha-mo everyday ballgame play-DUR.JD Lha-mo plays ball every day.
Tibet-LOC rice C maize etc. grow-DUR Do rice, maize and so on grow in Tibet？ Markers of the durative aspect may be put after an auxiliary verb of volitional ability. Without an auxiliary verb, however, the sentence may also represent some volitionality.
You-AG raise can-DUR Can you lift and carry this?
blo-bzang sick consequently attend class TAP come can-NEG-DUR blo-bzang is sick. So he's not able to come to class.
Distribution of these three markers of the Durative aspect is fairly distinctive, and to some extent, related to the context in some exceptional cases. Take sentence (31) for instance, "V+gi yod" is an expression used by the seller. A buyer will usually use "V+gi vdug". Besides, an interrogative expression may also bring some relational variation.

Merely-past aspect
The Merely-past aspect refers to the state when an action has just happened, or when an event has just finished. It is usually marked with "groo yin" or "groo red" in Lhasa Tibetan. The Merely-past aspect can be represented in writing by "groo yin" and "groo red"，as well as by "grabs yin" and "grabs red". In written Tibetan, "grabs" by itself means "near upon, drawn on" (adverb) ， or "preparation, prepare" (noun). So it is quite understandable for it to assume such a grammaticalized meaning when attached to a verb in the present tense. Besides, it is logically proper for "grabs" in the imminent aspect to be attached to the future tense of a verb. In addition，there is no negation in the merely past aspect--its negation in the sense of lasting time is expressed by the perfect aspect.
I bathe-STC-MER a moment later him-DAT phone back-PRO I've just finished bathing. I will phone him in a minute.

Now conference dismiss-MER they go-MER
The meeting is over，they've just gone.
Choice over the two markers of the Merely-past aspect can be made by whether there is an egocentric or non-egocentric narrative viewpoint. In most cases, the first-person pronoun refers to the narrator, but in some others (such as in performative sentences) the narrator can be inexplicit or omitted.
I plane take off-MER take a rest V need-PRO-M.W I just got off the plane, and therefore need some rest.
They mountain-ABL get down-MER you quickly C catch up-PRO They've just gone down the mountain. You may catch up if you speed up.
A sentence of the Merely-past aspect may have a volitional or non-volitional verb. However its subject takes no agentive marker even if it is a volitional verb sentence. Zhou Jiwen (2003:121) considered this to be a copular predicate structure, which needs no subject in the agentive case. This essay holds, however,that syntactically it is better to regarded "grab" as a marker of aspect.
He go to school-NEG-PRO he be ill-MER He's not going to school. He's sick.

I book read-MER
I have been reading a book.
The following two sentences taken from Jin Peng (1983:51) have different markers for the self-reference and the other-reference. They entail a different semantic meaning of "almost done". We are not to disscuss such phenomenon since there's no such case in our corpus.
I-AG medicine wrongly eat-MER I almost took the wrong medication.

He dress wash-MER
He almost washed the dress. (It shouldn"t be washed.) Both Zhou Jiwen (2003:122) and P. Denwood (1999:161) said that "verb+pa da ga + yin/red" can also mean "the action has just finished ". But Zhou took the first as in a sentence with a copular verb,while Denwood took it as an action of the perfect aspect.
I-AG meal cook-MER I have just cooked a meal. (Denwood) "pa da ga" is a common word in the Lhasa dialect principally functioning as a conjunction or an adverb. Its alternative spelling in spoken form is "pa dang" (M. C. Goldstein, 1978:668). Syntactically speaking, it is quite proper to this formal as a Merely-past aspect.

Realis
The Realis aspect in Lhasa Tibetan is marked by "pa yin" or "pa red"，with a preceding verb in the form of past tense usually illustrated in traditional grammar books. "Realis" means the action the verb denotes has either occurred, finished, or been realized. Hoshi, Michio (in Hu Tan, 2002), who said the Lhasa dialect has only two aspects, namely the perfect aspect and the imperfective aspect, considered "pa yin/pa red" and "song/byung" as markers of the Perfect-nondurative aspect. Jin Peng (1983: 52), however, took "pa yin/pa red" as markers of a tense (the Past tense) and "song/byung" as markers of the Perfect aspect. Qu Aitang et al. (1983: 80) considered "pa yin/pa red" to be markers of the Perfect aspect. Obviously, there are quite some differences not only in the terms but also in the concepts employed. ③ First of all，the notion of Realis has nothing to do with the notion of tense，for, as illustrated below, the aspectual marker "pa yin/pa red" can be added to all the verbs in the following sentences, regardless of whether the action (or event) is in the past, or at the present, or even in the future.
I school not come before Tibet-LOC work three years do-REA I had worked in Tibet for three years before coming to school.

People-GEN life good-COP change-REA
People"s life has been improved.
Example 47 is a statement of some past event or action, and therefore is in the past. Sentence 48, however, is a statement of some present events.
Secondly，Realis and Perfective are two different notions. "Realis" is about whether the action has been a fact; "Perfective" is about whether the action is finished. (Liu Xunning, 1988). Sentence 49 below is not concerned about whether the action of running away is completed or not, nor is there a notion of the end of being lost. All it tells is the reality that the children have got lost.
Flee-NOM-GEN on the way-LOC children these two get lost-REA The two children got lost while they were fleeing.
The two markers of the Realis aspect indicate two different sources of information, one with the speaker's self-knowledge, the other with inferred knowledge. Self-knowledge ③ In western linguistic literature, perfective-imperfective and realis-irrealis are syntactic categories of verbal aspect and modality respectively. This essay takes "pa yin /pa red" in Lhasa Tibetan as a syntactic marker of verbal aspect. I thank J. L. Randy for bringing up the topic of realis, which urged me to read and learn more about the issue in other languages and make a compartive study on realis. means the speaker is describing an event with his first-hand experience, and therefore with a subjective certainty; inferred knowledge means the speaker is speaking with his empirical deduction or is quoting from another information source. Further analyses show that the narrator's perspective is another important factor in the selection of Realis markers. Self-reference should be used to keep the agreement with the Ego, when the speaker speaks from his personal view (or an ego-centric perspective). With this provision，the narrator describes his first-hand experience with a first-person pronoun to express such a perspective. There is no need for any special marker of evidentiality. Non-egocentric actions, however, need some special formal marker to indicate the source of the speech.
Negative sentences in Realis are a bit different. Since the ending "pa yin /pa red" indicates the occurrence and realization of an action/event which can be either positive or negative, the negation of the realization of an event can only be achieved through the negation of the event. That is to say，the negative form of sentences in Realis must be expressed through the negation of the event. Consequently, the occurrence of an action or the emergence of an event can no longer be expressed with endings of Realis. Compare the following two sentences.

I-AG
you-OBJ joke-NEG-REA I'm not kidding you.
I-AG you-OBJ joke-REA I'm kidding you.
When a sentence has a Realis ending while negating an action or event, it usually entails some special meaning. As far as sentences with volitional verbs are concerned, this usually indicates a strong subjective opinion, meaning "the action has not been realized, though it could have been realized".
(52) khyed-tsho-i dzin-grwa-i mi As to non-volitional actions or events，whether they will occur or not is constrained by objective conditions. Since extraneous factor(s) can only be satisfied through the negation of an event, it usually entails some semantic emphasis on the objective cause.
that look for-TAP Go leg run break C NEG-find-REA Then the subjects looked for such foliage over all the country. They couldn't find it even when their legs were hurt.
The negation of Realis sentences ended with "pa red" is quite unique in that it is expressed through the direct negation of the existence of the event, with an "yog ma red" ending attached to the verb to mean the non-existence of the action. "yog ma red" is a grammatized negative form of an existential verb.
You now Tibetan food other taste-NEG-REA-M.W Oh，you haven"t tasted other Tibetan foods yet, have you?
Consequently word read-NOM that-AG in this way read-NEG-REA None the less, people don't read it this way.

Perfect aspect
Predicate verb markers of "song" and "byung" in Lhasa Tibetan are widely regarded as formal markers of the Perfect (aspect / tense), for example by C. A. Bell (1905:46), H. B.Hannah (1912, C. B. Shefts & Kun Chang (1981:303-321), Jin Peng (1983:52), Qu Aitang et al. (1983:82), and Hu Tan (2002. We think, however, that "verb in past tense+song/byung" indicates not just the occurrence of an action or the change of a state by the time of the speech, it also indicates the speaker"s observance of the change or new situation. None the less,the speaker"s emphasis is not on the completion but on the occurrence of the action. So we put it in the category of Perfect Aspect.
There are two formal markers of the Perfect apsect in Lhasa Tibetan，"byung" and "song". ④ Their distributions are commonly considered to be restricted by Person and ④ We will come to the issue of egocentricity of sentences with "song/byung". Here is just a brief illustration of the Perfect aspect using the "Ego" theory in Sociology. According to the sociologist G. H. Mead, there are two kinds of "self": the "subjective self" and the "objective self". The "subjective self" is revealed from experiences and is of the past; the "objective self" is happening and of the moment. The narrator's egocentric narration, therefore, is of events already happened in the past.
volitionality. Among the four combination patterns of pronoun and volitional verb, "song" can be used in three, and "byung" in two. Neither "song" nor "byung" appear in the "self-person volitional" verb pattern.
I childhood-ABL granddad-GEN horse-GEN age calculation method all study-PEF Granddad taught me how to assess the age of a horse when I was young.
He cough month one keep-STC-PEF Yin-nai na-tsha drags-ru gro-yas mi-dug. but disease recover-COP change-NOM NEG-have He has been coughing for one month now. There is still no sign of relief.
I colleg graduate you-OBJ surprise-PEF As a college graduate I am surprised to know that you are such a knowledgable man.
Girl this-GEN name I recall-PEF Now I remember the girl's name.
It is quite reasonable that "song" and "byung" don't appear in self-person volitional verb sentences. Firstly, sentences of this type indicate the speaker"s recent discovery of the occurrence of a stated action. "Discovery" means that the speaker was not personally involved in the event--he only got the information from personal observation--and any person involved in the event could only be referred to through a other-person pronoun. So this is knowing-by-sight evidentiality. On the other hand, if the speaker was involved in the event, then it was a non-volitional action of him (knowing-by-newly-found-information evidentiality). Secondly，there is a notion of distance-to-event-occurrence in the Tibetan aspect system, such as the contrast between the prospective aspect and imminent aspect (along with other differences), and the contrast between the perfect aspect and the merely-past aspect. The Realis aspect is used in Tibetan for events which happened earlier, taking up the "self-reference+volitional" combination. Nonetheless, the Perfect aspect may also suggest to a certain extent the occurrence of an event or the realization of an action. This was exactly the reason that scholars such as Denwood et al. considered "pa yin/pa red"，"song/byung" to be formal markers of one single category. In fact, there are some differences in terms of the narrative point of view.
The most outstanding feature of "song" and "byung" as markers for the Perfect aspect is that they entail a concept different from other concepts of aspect. The difference does not lie primarily in the pronoun used, or in the volitionality of the verb, nor in the temporal distance, nor in evidentiality. Let's look at some sentences first.
I sick when comrades-AG see-TAP times many come-PEF Comrades came to see me many times when I was sick.

I-GEN home-LOC stay-NOM guest
gone-STC-PEF yin-nai da-dung ca-lag mang-po bcol-byung. but still things many leave-PEF Although our guests have gone, they still left many of their belongs in our house.
These sentences with "byung" are all sentences of the self-person volitional verb pattern. A common feature is that all verbs direct to the speaker's cognitive center. A more detailed study will find that there are several types of cognitive centers: (a) the Ego of the speaker as center, as in sentence (63); (b) the location or position of the speaker as center, as in sentence (64); (c) the people/things that the speaker identifies with as the center, as in (65).
If a sentence is self-referent，it usually reveals what the speaker has personally sensed or noticed. In such circumstances, several different types of verbs are commonly found: "see, hear, know, understand, meet, feel sick, feel hungry, arrive" as uncontrollable actions; "get, remember, win, discover, dream, obtain" as unintentional actions. All of them are semantically self-referent, syntactically non-volitional, and result in either some new information related to the speaker, or something the speaker gets. i.e.，the speaker is the center of the consequence of the action or the direction of the action. For instance： (66) kho nag-chuu slebs-dug. Nga-s kho mgron-bsu-khang-la dzul-pa mthong-byung.
He nag-chuu come-RST I-AG he hostel-ALA entering-NOM see-PEF He's in Nag-chuu(county name) now. I saw him entering the hostel.
I-GEN pen black the lost-RST everywhere look for C found-NEG-PEF I have lost my black pen, for which I have looked all over but still haven't found.
Sentences with "song" and "byung" usually have verbs that are of the opposite direction, i.e.,actions moving from the center or not to the center, or results not unexpected or disliked by the speaker. Such verbs include "forget", "fear", "lose", "mistake", etc.
Parent-AG I-OBJ angry-PEF I-GEN pen that lose-PEF My parents are angry with me. I've lost my pen.
As we can see, the difference between "song" and "byung" relates closely with the underlying egocentric view of the speaker. Regardless of the pronoun used or the volitionality of the verb, the speaker will use "byung" for egocentric or self-beneficial expressions, but "song" for non-egocentric or non-beneficial expressions. As a result, volitional verbs can be used as non-subjective non-volitional verbs. For instance, sentence (70) indicates the speaker's nonvolitionality of having drunk a lot of alcohol by using the formal marker "song".
(70) a tsi de-ring nga-s chang mang-po zhe-po cig thung-song. The volitionality of the speaker is usually of prime importance. For instance, sentence (71) seems to be an action towards the self-referent speaker and therefore should have used "byung". But the speaker's intention is revealed through the use of "song" that "I" doesn't want to be recognized because of his certain personal interests. Similarly，"she" in (72) is not the central concern of the speaker, while the speaker still hopes that "she" would be recognized.
You she recognize-PEF-M.W she-PAP neighbor-GEN Wang aunt-GEN daughter is Have you recognized her? She's the daughter of Aunty Wang the next door.
Why is there the notion of the speaker's Ego? So far we believe that this notion has something to do with empirical knowledge about the world, the self including of our body, our behaviors and our thoughts. The main body versus object reflects what philosophers thought，knew. Consequently, man as the starting point of cognition enters a relationship with his surroundings in which he is the center of the relationship, i.e., the Ego. Though this cognition will gradually be improved as his relations with the physical world and the society, we can still observe this from a baby"s psychological processes. Such an egocentric viewpoint is deeply rooted in the knowledge system of man, and is revealed in certain languages. This is reflected in the self-direction we discussed earlier, and in features such as gaining or loosing according to the egocentric criteria influenced by the speaker's feelings or determination.
Why not use the "self-person volitional verb" pattern for the perfect aspect？This may have something to do with the social backgrounds of discursive construction. Since the realization of any discursive purpose depends on other people's reaction or attitude, direct expression of egoism (identity, interests) by any individual is restricted. Consequently, idiomatic expressions are developed that are apparently non-egocentric but actually ego-directed (as the other-referent sentences or self-referent non-volitional sentences in Tibetan). Such pragmatic or modality expressions may even be grammaticalized in time.

Resultative
The Resultative aspect bears the syntactic meaning that the occurrence of an action or the completion of an event has resulted in certain consequence. It has the formal marker "yod" or negative markers "med", "shag/mi dug", "yoo red/yoo ma red". Both Jin Peng (1983:54) and Denwood (1999:158) consider these to be formal markers of the Perfective. There are still quite a couple of disputes on whether Tibetan has the Perfective aspect. For instance, Jin Peng holds that besides the markers mentioned above, "song" and "byung" are also formal markers of the Perspective. What's the difference between these markers? The word "tshar" in Tibetan used after verbs does indicate the perfective aspect, but it is syntactically a complement.
We-AG visit-NOM-GEN agenda one make up-RST We"ve made up a timetable for the visit. (74) kha-sa gzugs-po bde-po-ma-byung tsang kha-lag ga-gai bzas med.
Yesterday body be well-NEG-PEF consequently meal any eat-NEG-RST He ate nothing yesterday because he didn't feel well.
Yesterday you-AG send a message-NOM the I-AG completely forget-RST I"ve forgotten your messsage completely. (76) kho-s las-ka yag-pa gcig byed-myong-yoo-ma-red.
He-AG thing good one do-STC-NEG-RST He has never done anything good.
The syntactic distribution patterns of the three markers of Resultative are fairly distinct. This may be due to the continued existence of a consequence, which helps people not to get confused. If an event is the narrator"s personal experience, the ending is "yod" , which may be taken as the archetype; if the event is seen-by-sight of the narrator but not of his personal experience, then "dug/shag" is used; if it is a personally experienced non-volitional action, "dug" may be used; and if the narrator is merely quoting hearsay or making experiential inference on an event he's not temporally/spatially present at, then "yoo red" is used.
I also deliver gift-NOM thing small such some buy need-RST.EG And I need to buy some small gifts.
Last year then rnam-grol Lhasa-LOC stay-RST.JD Rnam-grol was in Lhasa this time last year.
The Resultative aspect may also be used even if the event is yet to happen, as in (77). Volitional verbs used in non-self-conscious situations may be regarded as non-volitional actions, or as results of non-volitional actions (Sun, 1993), as in (79). Except for the cases in need of speculation,"yoo red" may also be used in circumstances long before or circumstances not present, as in (80).
With a color of resultative consequences to an event, markers for the Resultative aspect are usually used in sentences indicating the existence of an event or the consequence(s) of an action. For instance, sentences with typical verbs of result, verbs with resultative complement, or auxiliary verbs with a static tendency.
You-GEN dress that-GEN color completely fade-RST The color of your dress completely faded.
She-AG shout voice completely hoarse-RST She cried herself hoarse.

Contextual aspect
Although it is not possible for anyone to "witness" any event not yet happened, nor possible for a past event to be repeated, people are none the less able to make their prediction or estimation about such events. If such speculation is based on reliable experiences, then it is reasonable to make a firm judgment on the outcome of the event. Just as if they have "personally experienced or witnessed" the occurrence of an event. The syntactic meaning of a predicative verb indicates future result of an event, estimated outcomes of a completed event, or the speaker's certainty to such a result. So we call it the "Contextual aspect".
He-AG words that speak-NEG-RST He (definitely) wouldn"t have said that.
go Otherwise teacher Wang office-LOC wait for-CON Teacher Wang asked you two and me to see him as soon as possible. Let's go immediately, or (I'm sure that) Mr. Wang will be waiting for us in his office.
The formal ending for Tibetan verbs in Contextual aspect is "pa yod/pa dug". Denwood (1999:153) regards "pa yod/pa dug" as formal markers for the future tense,despite the fact that some sentences are inferences on the result of past events, as in (83). Zhou Jiwen (2003:134) regarded them as expressions of an experientially based speculation.
Having associated this marker with certain time structure and empirical knowledge of the speaker, we regard the combination as the "contextual aspect" so as to emphasize the empirical inference of the outcome of an event.
We-GEN house now-GEN size right suitable-CON The size of our present house is just right.
The difference between the Contextual and Resultative aspects is that the Resultative aspect is mainly a narration about the result of the event happened, while the Contextual aspect is an inference about the result of a future event. The Contextual differentiates itself from Prospective aspect in that the Prospective Aspect emphasizes the temporal relationship between events, while the Contextual aspect focuses on the outcome of the event. Let's compare (87) and (88) .
We two time how much arrive can-CON (You are sure ) in how long we can arrive?
Now that the Contextual aspect is an inference about a future event or an event already happened, it can only be used with the non-firsthand-experienced narrative mode. The experience can be used as direct evidence, however, since it has a high degree of certainty. Hence, this aspect usually employs the "pa dug" marker. Only sentences with a self-referent take "pa yod" among our examples.

Coda
Aspectual markers in Lhasa Tibetan are a complicated issue related with many syntactic categories. Particularly since egocentricity and evidentiality as pragmatic-semantic notions have been grammaticalized, they form a complicated multivariant expression system with other categories such as personal pronouns, reference, volitionality, mood, and so on. Apart from those discussed above, in Lhasa Tibetan there are a couple of other syntactic markers that may be combined for the expression of more complicated mood and modality We have composed a table of syntactic categories such as verbal aspects, evidentiality, egocentricity and others and their formal markers, so that the reader may have a clearer overall picture on these categories. This table is attached to the end of the paper as Appendix II.