I suddenly had a day to myself, so after perusing my books I found on my shelf, I took a quick look at the web. The material on the perfect in Latin is puerile, absolutely no linguistic sophistication, just a bunch a tired old tropes from a mountain of high school grammar books. Useless. If anyone knows of a real treatment of tense and aspect in Latin on the web, please let me know.
The surest thing is to go to the corpus. Several things came up though just from the books I checked. I don’t think I found anything earth-shaking, but I had just never read or heard (I took a class in The History of Spanish) anything about a major change in the function of aspect form, yes, but not function. It seemed odd to me that Latin would use the imperfect for a simple past statement like “Yesterday was Wednesday” and Spanish, deriving straight from Latin, would use the perfective form (preterit).
So here goes:
Comrie, Elcock, and Palmer all state that the Latin perfect form is a merger of the I-E aorist and the perfect.
[time out for definitions: perfective is the aspect of action begun or completed, whether in the past or not (Latin has this only in the past whereas Russian has it in both the past and the future). Perfect marks a state resulting from an action; both English and Spanish separate these two whereas Latin combined them in the Present Perfect. Aorist = an event as an item of history regardless of its duration.]
Palmer (The Latin Language) says the Latin perfect marked a state resulting from an action and also denoted that an action is over and done with. The present perfect can be used of repeated actions e.g. saepe dixi, just like we have in Spanish when we say Cada dia fui a la playa, using the preterite = perfective because you’re summing up the action of each day.
Elcock (The Romance Languages) says “the range of tenses with past aspect considerably exceeding that of Classical Latin.”
Neither of these two scholars of Latin and the daughter languages indicate any shift in how aspect is used.
So I turned to references to Latin in two standard books on tense and aspect, both by Bernard Comrie and titled with their topic.
Comrie distinguishes between perfective and perfect, the former being an action limited and the latter marking a state resulting from an action.
Comrie states that the Latin perfect shows its two components: preterit and perfect, by pointing to the sequence of tenses and how the Lt. Perfect can be either a primary tense if it functions as a preterit or secondary if it functions as a perfect. He states that there is no reason to treat the Lt. Perfect as having 2 covert verb categories; it’s a matter of interpretation, or, as we say, of context, which one is meant. He says the Lt. perfect operates like the French in merging the 2 functions, a non-perfect past and a perfect, but French uses an earlier perfect form (j’ai chante’) whereas the Latin uses a non-perfect past form for the perfect as well. It is no wonder, to me, that school grammar books gloss over this by simply saying the Lt. Perfect can be used to “mean” either I X’d or I have X’d; this clearly comes out of the tradition of translating as a way of getting to meaning. Keller & Russell state on p. 84 “The perfect is the only tense of the indicative that hs two different times with two different aspects (see paragraph 6). It is therefore important always to keep in mind two translations. For example rexi may be translated “I rules (past time, simple aspect) or “I have ruled” (present time, completed aspect). Context helps to determine which translation is correct.” I picked on Keller & Russell b/c I think they do better than most when you go to paragraph 6. p. 29, and read: “Perfect: a. reports an action in present time with a completed aspect (she has thought) or b. reports an action in past time with simple aspect. (she thought)” See below * for my problem with “completed”, but they use the word ‘aspect’, a real step up.
It is interesting that Spanish is the only Romance language to split the past into a perfect and a non-perfect past e.g. cante’ vs he cantado, in the spoken language. So we can see by that that the tense/aspect system of Latin did undergo changes, but aspect is such a deep and universal category (note how Creole languages create aspectual distinctions without tense distinctions), it would be a wonder to me to see any change to that part of the system of meaning.
The descriptions of the imperfect in the linguistic texts, in the linguistic treatments of the Romance languages and even in the school grammar books all back up the concept of the imperfect as describing an action in the past without limiting it or defining it, just as a state or on-going action. Habitual and repeated aspects also frequently derive from this as do some others like “conative” (= try to). It is clear from I-E morphology that the perfective aspect is primary whereas the imperfective is innovated in the various branches. Comrie states on p. 72 of Tense, “If we take it that it is most natural for a past tense verb to have perfective meaning, the it is natural for a language to seek some other means of expressing a past tense that does not indicate a single complete action, and it is here that the Imperfect/Aorist distinction enters. In fact, the Imperfect expresses in past tense an aspectual value that is more typical of the present. In traditional I-E linguistics, the Imperfect is often characterised as the ‘Present in the Past’, which captures the above observation that the Imperfect expresses a typically present tense aspectual value in the past tense.”
*Let me close by arguing for a much clearer understanding of the perfective aspect. I rely on William Bull for this basic concept. When we talk about an action or state in the past, we can divide them into two kinds: cyclic and non-cyclic. A cyclic verb happens only when it is completed (thus the erroneous focus on “completed action” of most school grammars) and to repeat the action, one must start from the beginning e.g. “I closed the door” cannot be said until the door has shut; to close the door again, one must open it, “undo” the action of closing. Thus the term cyclic as the action goes in cycles if it is repeated. Similar actions are “He dropped his keys.” “They crashed the car.”
Non-cyclic verbs, OTOH, happen once they start; all they have to do is start and you can say they occurred. There is no focus on the end or “completion” of the action. Thus: “The dog barked”. The focus is on the start of the act of barking with no reference to its end or completion. The dog does have to stop barking before you can say he barked; the end point is immaterial. The action is repeated or is simply on-going without cessation (as oft happens with barking dogs). Similar actions are “The boy ran.” “The teacher spoke.”
The BIG MOMENT is when I point out that BOTH require the perfective form, which in Latin is the Perfect Tense.
The opposite of the perfective is the imperfective, which serves to put the listener/reader in the middle of the action in the past. “The boys were running.” “The teacher was speaking.” He was dropping his keys (???????)”. “They were crashing the car.” (??????) The ? means these forms are unlikely to occur in normal language; you have to invent special conditions like filming a car crash scene for a movie where even a seemingly instantaneous action can be interrupted as in “They were crashing the car at just the moment the starlet walked by and fatally distracted them.” Weird, but possible. More likely “in the middle of an action” cyclic verbs are “He was closing the door when he heard a moan.”
For English speakers, the real problem is the way the simple past form can be used to express both perfective and imperfective aspect, as in “He shot duck”, perfective if the context is, “George went hunting with his dad and he shot duck” and imperfective if it is, “George shot duck as a kid going out with his dad.”
In English, it gets complicated with imperfective forms like “he would shoot duck as a kid” and “he used to shoot duck as a kid” and so on. So “Yesterday was Wednesday” in English gives us no clue as to what “tense” (aspect) would be used in another language b/c “was” functions in either aspect and has further complications as a linking verb. However, the argument that “erat” would be used in Latin simply b/c it occurs more often makes no sense to me. That’s like saying, “-ed is the past tense form occurring most often, so the past tense of “sing” must be “singed”. No; it’s whatever is appropriate and grammatically accurate for the language. In this case, I cannot justify the use of an imperfective form, “erat”, in this linking verb function in a simple past, the non-perfect past, so-called. The state of being Wednesday is limited and defined in terms of time just like “I drank four bottles of beer” or “I drank beer last night” are limited and defined by the number four or the specific time frame and would also require the perfective vs “I drank beer”, which would require the imperfective form normally unless there was a broader context suggesting a limit of some sort.