In this part, I will offer a short history of the concept of tense. In Part 3 tense usage in tensed languages and Bikol as a tenseless language, and in Part 4, my view of tense and my conlang’s tense features.
RECENT HISTORY of the CONCEPT of TENSE
Otto Jespersen
Jespersen started the truly scientific study of tense by not making the languages he describe conform with the Latin tense paradigm like what most of his predecessors and contemporaries did but instead he explicitly distinguished the category’s form and function with its notional meaning, like how he cobbled together “”the seemingly contradictory multiple past theory, Varronian tense-aspect theory and relative tense theories” [1▼]. One big advance he made is his distinction between categories (e.g. preterit ) and notions (e.g. past time semantics), where a list of notions is given for each category. His categories are language-independent (thus can be used to describe different languages) and his notions may be distributed differently among the different categories in different languages. “He recognized that languages differ in their number of tenses without necessarily failing to express the same times. His functional categories are options languages can select from; the tense system of each language is a partial exemplification of one universal system. The notions are distributed among different categories.” [2▼]
Jespersen, as far as I know, is the first to point out that English has no future tense. To him, English futurity is not expressed at all on the verb, or if expressed, by auxilliaries that do not just signify mere futurity but something else besides it that are not totally obliterated by history: volition (he will start at six), destiny (the congress is to be held next year) uncertainty (he may come yet) or obligation (I shall write to him tomorrow) [3▼]. For a further discussion why English has no future tense, please check out this post.
The start of Jespersen’s tense discussion is Madvig’s Latin tense description, which I show here with Madvig’s examples and Jespersens notations and additional labels.
|
|
I |
II |
III |
|
|
praesens |
praeteritum |
futurum |
| 1 |
in praesenti |
scribo |
scripsi |
scribam |
| 2 |
in praeterito |
scribebam |
scripseram |
scripturus eram (fui) |
| 3 |
in futuro |
scribam |
scripsero |
scripturus ero |
The drawbacks he found in the above model are: (a) scribam is found in 2 places: praesens in futuro (I3) and futurum in praesenti (III1) while the rest of the other forms are shown only once, and (b) praesens in praeterito (I2, “scribebam”) and praeteritum in preasenti (II1, “scripsi”) “are not indicated with sufficient precision by their places in the system, as shown incidentally by Madvig’s placing scripturus eram and scripturus fui at one and the same place (III2). These two are not synonymous, being distinguished exactly in the same way as scribebam and scripsi, but the distinctions, to which we shall have to revert, has really nothing directly to do with the other time-distrinctions contained in the scheme” [4▼].
Since he thinks that the above scheme is purely logical and symmetrical without regard to the way those 9 categories are actually used and represented in these languages, he reduced it from 9 (3×3 scheme) to 7 tenses by merging I2 with II1 and I3 with III1, then arranged them left to right in a time line, getting the insight from the tense names he assigned previously: before-past > past > after-past > present > before-future > future > after-future. He thought that this is an improvement over Madvig’s by avoiding its errors : (a) the “tripartition of “now”, which as a point has no dimensions and cannot be divided”(p.256), and (b) the arrangment of “time in a 2 dimensional scheme with 3×3 compartments. For there can be no doubt that we are obliged (by the essence of time itself, or at any rate by a necessity of our thinking) to figure to ourselves time as having one dimention only, thus capable of being represented by one straight line.” (p.256)
It is quite clear that Jespersen assumed that tense has the notion of one dimentional time, and since time is linear, then tenses can only be modeled in a one-dimentional, linear scheme, with Present as a point and not a span of time. Having first recognized 7 tenses, then divided the timeline among them and made a one-to-one relation between time and tense, he came up with the following:
| Tense (Grammatical) |
Time Divisions (Notional) |
Example |
| Ante-preterit |
Before-past |
had gone |
| Preterit |
Past |
went |
| Post-preterit |
After-past |
should/would go |
| Present |
Present |
go |
| Ante-future |
Before-future |
shall/will have gone |
| Future |
Future |
shall go |
| Post-future |
After-future |
? |
His term “Before” in Before-past is taken from Old English where such tense is constructed with the adverb “before”. In each of these divisions, there are retrospective and prospective versions, which are the perspective from which an event is viewed on the timeline. The prospective looks forward into the future, the restrospective into the past. Prospectives can be indicated by inserting expressions like “on the point of, about to or going to” (She is about to cry).
A minor weakness of his scheme is the presence of tense distinction Post-future which he said served just as a theoretical curiosity. A major weakness of his model is the lack of proper treatment accorded to relative tenses, due to 2 reasons:
- His assumption that tense is the gramatical representation of the notion of a one dimensional time prevented him from recognizing that retrospective tenses (perfect or ante) can be viewed simply as pasts relative to the main tense divisions (past, present, future), and the same for prospective tenses (conditionals or post), as Madvig had treated the relative tenses of Latin.
- His assumption that Present is a point without duration, thus he has difficulty disambiguating the Present Perfect as Present (his view) from Present Perfect as Ante-present (from symmetry with other perfects), for “the perfect cannot be fitted into the simple series, because besides the purely temporal element, it contains the element of result. It is a present, but a permansive present: a present state as the outcome of past events, and may therefore be called a retrospective variety of the present. That it is a variety of the present and not of the past is seen by the fact that the adverb now can stand with it: “Now I have eaten enough” (p. 269). It seems that to him, for it to be fitted into the scheme, the Present Perfect, if an Ante-present, must occur with a past temporal adverb (“yesterday”), for it to be prior to a punctual present time. This resulted in his model leaving out Ante-present or Present Perfect (“has gone”). He said that “the present time..is a point, which has no duration…The present moment, “now”, is nothing but the ever-fleeting boundary between the past and the future…But in practice “now” means a time with an appreciable duration, the length of which varies according to circumstances.” [5▼]
Hans Reichenbach
Reichenbach [6▼], after considering the tenses of a few illustrative passages, saw “that we need three time points even for distinctions of tenses which, in a superficial consideration, seem to concern only two time points”, and knew why Jespersen had difficulty distinguishing the present perfect from the simple past: for although Jespersen indicated the 3-point structure for past perfect and future perfect, he did not do the same for the present perfect and other tenses.
He analyzed tense using 3 primitives: S (point of speech), E (point of event) and R (point of reference) and come up with 13 possibilities of ordering the time points. R’s position relative to S indicates past (R_S), present (R,S), future (S_R), and E’s position relative to R indicates anterior (E_R), simple (E,R), and posterior (R_E). He said that since R vs S and E vs R relations yields 9 tenses and since E’s position relative to S is irrelevant, we can call these 9 tenses as fundamental. He represented the English tenses as follows (comma represents simultaneity, dash or underscore means time interval):
| Traditional Name |
Reichenbach’s Name |
Point Relation |
Example |
| Simple Past |
Simple Past |
E,R_S |
I saw John |
| Present |
Simple Present |
S,R,E |
I see John |
| Simple Future |
Simple Future |
S_R,E |
I shall see John (tomorrow) |
| Past Perfect |
Anterior Past |
E_R_S |
I had seen John |
| Present Perfect |
Anterior Present |
E_S,R |
I have seen John |
| Future Perfect |
Anterior Future |
S_E_R S,E_R E_S_R |
I shall have seen John |
| – |
Posterior Past |
R_E_S R_S,E R_S_E |
He would do, or I would see John |
| Simple Future |
Posterior Present |
S,R_E |
(Now) I shall see John |
| – |
Posterior Future |
S_R_E |
I shall be going to see him |
Of the 9 fundamental tenses, he said 6 are possible in English grammar, for (a) two tenses are realized by one construction, i.e., simple future is ambiguous between S,R_E and S_R,E, (b) the Posterior Future has no established form in English but can be expressed as indicated above, and (c) the Posterior Past is not officially recognized as a tense in English, but sometimes classified as a tense of the conditional mood. But countered this by writing that its not a conditional usage, being derived from simple future by backshifting R and E.
Of the perfects and future tense origin, Reichenbach mentioned that “‘I shall go’ meant originally ‘I am obliged to go’; the future tense meaning developed because what I am obliged to do will bee done by me at a later time..In Old English no future tense existed, and the present tense was used both for the expression of the present and future. The word “shall” was used only in the meaning of obligation. In Middle English, the word shall gradually assumed the function of expressing the future tense.” As for the perfect, “the double function of ‘have’, as expressing possession and a past tense, is derived from the idea that what I possess is acquired in the past; thus ‘I have seen’ meant originally ‘I possess now the results of seeing’, and then was interpreted as a reference to a past event… This is even more apparent when a two-place function is used. Thus ‘I have finished my work’ means originally ‘I have my work finished’,i.e., ‘I possess my work as a finished one’.”
Sten Vikner
Vikner [7▼] reworked Reichenbach tense system, criticizing Reichenbach as (1) there are no linguisitic evidence for the identical forms in Simple Future and Posterior Present, (2) there’s lack of linguistic evidence for Posterior Future in English, Danish and French, as it deviates from the others in morphosyntax, (3) there’s no putative Future Perfect of the Past tense in Reichenbach tense system and would be problematic in describing it in Reichenbach system, and (4) the inadequacy to differentiate the three possibilities of future perfects (S_E_R, S,E_R, E_S_R) generated by Reichenbach system, as in his own words “no way of knowing (from the tense form on its own) where the event point occurs in relation to the speech point (before, at the same time, or after)”. He distilled his tenses into 3 features ±past, ±future, ±perfect shown as follows:
| Feature |
Relation |
Morphosyntax |
| +past |
1_S |
the first element ends in the morpheme -ed |
| -past |
S,1 |
the first element ends in the morpheme -s or Ø |
| +future |
1_2 |
the first element is a form of will/shall |
| -future |
1,2 |
absence of will/shall |
| +perfect |
E_2 |
the penultimate element is a form of have |
| -perfect |
2,E |
absence of form of have |
resulting into 8 tenses, with the following specifications:
| Tense |
Past |
Future |
Perfect |
| Past |
+ |
- |
- |
| Present |
- |
- |
- |
| Future |
- |
+ |
- |
| Future of Past (Conditional) |
+ |
+ |
- |
| Past perfect |
+ |
- |
+ |
| Present Perfect |
- |
- |
+ |
| Future Perfect |
- |
+ |
+ |
| Future Perfect of Past (Conditional Perfect) |
+ |
+ |
+ |
He introduced his own symbolization using two reference points (1 for R1 and 2 for R2), his maximum in a sentence (“the maximum number of non-coreferent time adverbials possible in a sentence”).
| Tense |
Notation |
Example |
| Past |
1_S / 1,2 / E,2 |
He worked. |
| Present |
1,S / 1,2 / E,2 |
He works. |
| Future |
1,S / 1_2 / E_2 |
He will work. |
| Future of Past |
1_S / 1_2 / E,2 |
He would work. |
| Past perfect |
1_S / 1,2 / E_2 |
He had worked. |
| Present Perfect |
1,S / 1,2 / E_2 |
He has worked. |
| Future Perfect |
1,S / 1_2 / E_2 |
He will have worked. |
| Future Perfect of Past |
1_S / 1_2 / E_2 |
He would have worked. |
He said “This symbolization stresses the hierarchical nature of the relations between S, R and E: S is the most independent, R can only be placed in relation to S, and E is even more dependent, as it can only be placed in relation to R which in turn is dependent on S…Nothing can be said of the position of E relative to S, as these two points cannot be related to each other. E can be related to R, not to S..” and revising further “As one reference point, R, has been discarded in favour of two, R1 and R2, and .. none of the reference points..are both directly related to S and following S in time..”
He said the above relations “suggest a connection between markedness and coincidence between points: In any of the three relations, coincidence is the unmarked option, as the tense intuitively felt to be the least marked, present, has coincidence in all three relations, whereas the most marked tense, future perfect of the past, has no coincidences at all”. He noted that the “three two-place relations approach results in a systematically very strict (and non-overgenerating) account,.. [with] interesting consequences for… time adverbials”, and “allows description of the rule governing the choice of tense in subordinate clauses introduced by certain temporal conjunctions.” Vikner only claims his system to be valid for English, French and Danish but suspect its validity to be universal. He was not the pioneer though in coming up with three binary relations, but Lammert Te Winkel who worked with Dutch in 1866, another Western Germanic language like English and Danish, and got past vs present, synchronous vs posterior, and imperfect vs perfect.
Other Authors
Other works on tenses will be added once I get access to them:
1. William Bull (1960) Time, Tense, and the Verb. University of California Publications in Linguistics, Vol 19. Berkeley: University of California Press. – discusses about 12 possible tenses.
2. R Declerk, From Reichenbach(1947) to Comrie (1985) and beyond” Lingua 70:305-364, 1986
3. Comrie, Tense.
(to be expanded)
Notes:
[1▲] Robert Binnick, Time and the Verb, p.54
[2▲] Robert Binnick, Time and the Verb, p.58
[3▲] Otto Jespersen, The philosophy of grammar, 1924, p. 50
[4▲] Otto Jespersen, The philosophy of grammar, 1924, p. 255
[5▲] Otto Jespersen, The philosophy of grammar, 1924, p. 258
[6▲] Hans Reichenbach, The Tenses of Verbs, section 51 of Elements of Symbolic Logic (New York: The Macmillan Company, 1947) 287-98, 1947.
[7▲] The philosophy of grammar, Sten Vikner, Reichenbach revisited: One, Two, or Three Temporal Relations?
1985
Tahoŋ kadlagan чakoh (I am a man of the forest).
