Development of the Catalan Periphrastic Past Tense Construction

Despite the name, Indo-European Syntax at NYU is much more of a historical linguistics class than a syntax class, reflecting the expertise of Professor John Costello.

This essay is a review of the development of Catalan's unusual periphrastic past tense form. It's a little denser than most of the other posts on this site and assumes some prior knowledge of syntax and morphology. If you have any questions, please ask.


In the Catalan language, there are three important periphrastic constructions that can be formed with the auxiliary verb anar (“to go”). One is the anar + present participle construction, which is functionally and formally equivalent to the Spanish ir (“to go”) + present participle construction. This construction has the same tense as the tense in which the auxiliary anar is conjugated, and has a progressive aspect. Another construction is the anar a + infinitive construction, which is formally equivalent to the Spanish ir a + infinitive construction, but has a slightly different function: instead of functioning as a periphrastic future tense, as in Spanish, the Catalan construction carries the same tense as the conjugated anar and has an inchoative aspect. The third periphrastic construction with anar is the one that interests us here: it is formed with a non-standard present tense conjugation of anar followed by the infinitive, and functions as a periphrastic past tense. In this essay, I will provide a brief overview of the characteristics of this construction that make it unusual from a linguistic perspective, and outline Antoni Badia’s argument of the development of the periphrastic past tense.

The periphrastic past tense construction is of interest to us for several reasons. Firstly, the conjugation used is not the standard present tense conjugation of anar. Anar in the present tense normally conjugates as follows:

Standard present tense conjugation of anar

1st person singular (1s)vaig1st person plural (1pl)anem
2nd person singular (2s)vas2nd person plural (2pl)aneu
3rd person singular (3s)va3rd person plural (3pl)van

The irregular conjugation is due to syncretism of the Latin verbs amnare and vadere as they evolved into Catalan; an ir- root also appears in the future and conditional conjugations of standard modern Catalan, derived from Latin ire. (“anar”, Diccionari) This syncretism can also be found in many of the Western Romance languages, such as Spanish and Portuguese ir (syncretism of Latin ire, vadere and the perfect forms of esse) and Italian andare (syncretism of Latin amnare and vadere). However, when anar is used as part of the periphrastic past tense, it conjugates as follows:

Present tense conjugation of anar in periphrastic past tense

1st person singular (1s)vaig1st person plural (1pl)vam
2nd person singular (2s)vas2nd person plural (2pl)vau
3rd person singular (3s)va3rd person plural (3pl)van

Besides the non-standard conjugation, this periphrastic construction is also significant because of its function. Other Western Romance languages have a formally similar construction that is functionally distinct. In French, for example, the aller (“to go”) + infinitive construction is formally equivalent to the Catalan periphrastic past, but this construction in French serves instead as a periphrastic future tense, rather than as a past tense. The same happens in Portuguese with a periphrastic ir (“to go”) + infinitive future tense construction. As mentioned previously, Spanish has an ir a + infinitive construction that functions as a future tense. Among the major Western Romance languages, Catalan is the only language to form a periphrastic past tense with a present tense form of the auxiliary verb “to go”.

Historically, the preterite form was preferred to the periphrastic form. However, in the written language, the periphrastic past tense now occurs in free variation with the simple past (preterite), and in the spoken language, the periphrastic past tense has virtually replaced the simple past, except in the Balearic Islands, where the simple past continues to be used in speech. Grammars of Catalan are very emphatic in saying that the periphrastic past and the simple past are semantically identical, even if one form may be preferred over the other for stylistic or historical reasons. Antoni Badia, for example, says that “today literary Catalan employs both tenses equally, such that the choice has become a stylistic device… It need not even be said that meaning of the two perfect forms is always identical” (Gramática Catalana, 277). Alarcos Llorach, in his Estudis de Lingüística Catalana, mentions several times while analyzing the preterite form portí (“I carry”, from portar, “to carry”) that it is equivalent to vaig portar (125, 130, 132).

Another reason the periphrastic past tense construction is of interest to us is that the Catalan periphrastic past tense has undergone formal changes since its first appearance in writing. In the earliest written texts such as the Libre dels feyts del rey en Jacme, also known as the Crònica del Jaume I (written in the 13th or 14th century), it appears that the periphrastic past tense might not have been formed with the present tense of anar, but with the preterite conjugation. In Gramàtica Històrica Catalana, Badia points out that the phrases van ferir (with the auxiliary in the present tense, third person plural) and anà’l ferir (with the auxiliary in the preterite, third person singular) both appear in the Crònica. He argues that since the Provençal and Castilian of that era had periphrastic present tense anar + infinitive and ir + infinitive constructions respectively, it seems likely that van ferir was a periphrastic construction for the present tense, and the periphrastic past tense construction could instead have been anà’l ferir. (370) This suggests that at some point, while the Spanish and Provençal forms involving the conjugated present tense either fell out of use or, in the case of Spanish, evolved to become a periphrastic future tense, in Catalan the periphrastic construction evolved in the other direction to become a periphrastic past tense. How did this evolution happen in Catalan, given that it did not happen in any other Western Romance language?

Badia has reconstructed, using “traces that are found in the texts”, a possible early form of the conjugation of anar as used in the periphrastic present and past tenses (Gramàtica Històrica Catalana 370):

Possible historical present tense conjugation of anar

1st person singular (1s)vau1st person plural (1pl)anam
2nd person singular (2s)vas2nd person plural (2pl)anats
3rd person singular (3s)va / vai3rd person plural (3pl)van

Possible historical preterite conjugation of anar

1st person singular (1s)aní / ané1st person plural (1pl)anam
2nd person singular (2s)anist2nd person plural (2pl)anats
3rd person singular (3s)anà3rd person plural (3pl)anaren

Badia suggests that as the preterite form became replaced by the periphrastic form, the fact that the 1pl and 2pl forms of anar in the present tense and simple past were identical, coupled with the “constant mix of the historical present and the perfect in old texts”, caused the present tense to be substituted for the preterite in the auxiliary verb anar, resulting in a conjugation that corresponds to the one still used in the modern Algherese dialect of Catalan (Gramática Històrica 371):

Present tense conjugation of anar in modern Algherese, used in the periphrastic past tense

1st person singular (1s)vaig1st person plural (1pl)anam
2nd person singular (2s)vas2nd person plural (2pl)anats
3rd person singular (3s)va3rd person plural (3pl)van

Badia postulates that once the periphrastic past tense became a definitive substitute for the simple past tense, Catalan speakers began to treat, for example, va venir (“he came”) as a single unit, almost as if it were a synthetic form of the verb with an inflected prefix (Gramàtica Històrica 371). Thus, he argues that the 1pl and 2pl forms became regularized to vam and vau to match the other conjugated forms, resulting in the modern standard Catalan conjugation of vaig, vas, va, vam, vau and van (Gramàtica Històrica 371). Because the regularization was tied specifically to the use of anar as an auxiliary, this did not affect the standard conjugation, creating the distinction we see today in the two different conjugations of anar.

There is an alternative conjugation, used only in the periphrastic anar + infinitive construction, that occurs in free variation in Central Catalan with the conjugation shown above in the first verb table:

Alternative present tense conjugation of anar in periphrastic past tense

1st person singular (1s)vàreig1st person plural (1pl)vàrem
2nd person singular (2s)vàres2nd person plural (2pl)vàreu
3rd person singular (3s)va3rd person plural (3pl)varen

The variant forms vàreig, vàres, vàrem, vàreu and varen, Badia argues, came about because if the present tense 3pl conjugation van were a regular conjugation, this would yield a preterite 3pl conjugation of varen, and consequently a var- root propagated into the 1s, 2s, 1pl and 2pl forms with varying degrees of incidence (Gramàtica Històrica 371). This argument seems plausible because if varen were indeed the preterite 3pl form of a verb var, the preterite 3s form of the verb would remain va (compare preterite 3s conjugation cantà for the verb cantar). Badia also notes that of the var- forms of the verb, the 1s vàreig is the least commonly used, which lines up with his hypothesis of a regularized var- form; in this conjugation, vàreig would be an irregular form (compare preterite 1s conjugation cantí for cantar) and it makes sense that it occurs less often than the others. (Gramàtica Històrica 371)

There remains much to be studied in the Catalan periphrastic past tense construction. Given that there are two different conjugations for the same periphrasis in free variation, and the periphrastic form itself is still in free variation with the synthetic form of the verb, at least in the written language, it seems likely that a shift towards one form over the other two will eventually occur, although it will be a slow process. Moreover, it would be instructive to study similar processes of verbal periphrasis in closely-related languages, particularly Occitan, to see if there were any unique linguistic characteristics that caused this periphrasis to develop in Catalan but not in any of its neighboring languages.


Works Cited

  • “anar.” Diccionari català-valencià-balear. Palma de Mallorca: Editorial Moll, 1980-3. Print.
  • Badia i Margarit, Antoni M. Gramática Catalana. Madrid: Editorial Gredos, 1980. Print.
  • Badia i Margarit, Antoni M. Gramàtica Històrica Catalana. Valencia: Biblioteca d’estudis i investigacions, 1981. Print.
  • Llorach, Alarcos. Estudis de lingüística catalana. Barcelona, Editorial Ariel, 1983. Print.

"Want" as Future Auxiliary in Singlish

Here's a little something I noticed a while ago. Singlish allows you to do something interesting with the verb "want", but only under certain circumstances.

For those who are familiar with German or with the history of the English language, you will notice that what I am about to outline in Singlish is a parallel phenomenon to what happened in English. German "will" means English "want" (verb). Formerly, in English, "will" did use to mean "want", and you can see this sense of the word in "last will and testament", or "I willed it into existence", or "where there's a will, there's a way." Now, of course, we primarily use "will" in English as a future auxiliary, to indicate that an action will be happening in the future.

This particular semantic drift most likely occurred because if you want something, it is in the future, not in the present. (As for why something similar didn't happen in German: language change is explainable but not predictable.)

I confess, the first time I was given this explanation, part of me refused to accept it. Of course my brain understood why it was likely, but the fact that "will" is already grammaticalised as a future auxiliary blinded me to the process - I could not conceive of making that leap from "will" as in "want" to "will" as in future tense. I took the explanation at face value but didn't like it.

So imagine how intrigued I was when, while working on a documentary, this gem of a Singlish sentence showed up in the Singaporean musical I was following:

"I eat until I want to bao zha (explode) already!"

Any Singlish speaker will recognise this as a valid construction. "Want" here, however, does not indicate volition. You cannot substitute "would like" in the sentence: "I would like to explode." Neither can you substitute "will": "I eat until I will explode." The only acceptable substitution is the "going to" future: "I eat until I [am] going to explode already."

The first explanation is that Singlish acquired this construction from Chinese, and you can see it in a direct translation:

"I eat until I want to explode already!"
我吃到我要爆炸了!

The question, then, is whether Chinese uses this construction to mark the future. The answer is: it does. Sort of.

As with many things in Chinese, the exact meaning is context-dependent, but take me at my word for now:

我要上学了。
Lit: I want to go to school already.
Meaning: I'm going to go to school now.

我们要吃饭了。
Lit: We want to eat already.
Meaning: We are going to eat now.

他们要走了。
Lit: They want to go already.
Meaning: They are going to go now.

Other meanings are possible, but I don't want to get too much into the complexities here. The important thing is, in all three cases, a Singlish speaker can hear the literal translation and accept the dynamic translation as meaning the same thing.

Without actually conducting full-scale research on this, my guess is that it is the "already" inchoative aspect marker that does the trick - when used together with "already", "want" is focused from a generalised volition down to a specific, immediate-future time frame. Take away the "already" or the 了, and suddenly volition becomes the preferred interpretation in almost any context:

我要上学。
Lit: I want to go to school.
Meaning: I want to go to school.

我们要吃饭。
Lit: We want to eat.
Meaning: We want to eat.

他们要走。
Lit: They want to go.
Meaning: They want to go.

Curiously, not all verbs can be used in all meanings with this construction. Not to be deliberately crude or morbid, but I think Singlish speakers will recognise these:

"I work until I want to die already."
"I want to vomit blood already."

The volition interpretation is not possible in the above sentences, in the same way it is not possible to willingly want or choose to explode.

The thing is, of course, when the verb has a neutral or positive connotation it is much easier to assume that volition is intended. For example, you would have no issues with the usual English definition of "want" in these sentences:

"I want to sleep already."
"I tired until I want to sleep already."

In fact, in the second sentence above, it is not possible to substitute "want" with the "going to" future. Volition is pretty much the only interpretation that makes sense.

The answer to untangling this construction lies somewhere in Chinese grammar. That's as far as I've worked it out in my head.

“Already” / “Liao” as aspect marker in Singlish

This is an essay I wrote for LING-UA 36 Indo-European Syntax at New York University.

Despite the name, Indo-European Syntax at NYU is much more of a historical linguistics class than a syntax class, reflecting the expertise of Professor John Costello.

Even though Singlish appears to have nothing to do with Proto-Indo-European syntax, this oddball topic was accepted as an essay for this class, because it studies a change in syntax resulting from language contact (another one of Professor Costello's research interests), in which one of the languages in contact is Indo-European.

This essay is not a research essay, but rather an outline of ideas. As a result, it does not meet the burden of proof required of a proper research paper; temper your expectations accordingly.

To the best of my knowledge, the syntax of this aspect marker has never been described in this way, although if that is not the case, please let me know.


Singlish is an English-based creole language spoken in Singapore, influenced by various Chinese dialects, particularly Hokkien, Teochew and Mandarin, Malay, and to a lesser degree, languages from the Indian subcontinent. It is often thought of as "English with Chinese syntax", although as we will see in this essay, this is not strictly true. In this essay, I will analyze the function of the word "already", as well as the closely related “liao”, as it is used in Singlish.

“Liao” is the Hokkien cognate of Mandarin Chinese 了 (lè), which has two primary functions in Mandarin Chinese: it can mark completion, thus functioning as a perfective aspect marker, or it can indicate a change of state, functioning as an inceptive or inchoative aspect marker (Ross and Sheng Ma 62, 226, 236). “Liao” in Singlish has these two functions as well. In Singlish, “already” is used in the same way as “liao”; “liao” can always be replaced with “already” except in one case, and the reverse substitution of “liao” for “already” is not always valid, as we will see in this essay.

A common and unambiguous use of Singlish "already" or “liao” can be seen in the following sentence: "Eat already or not?" / “Eat liao or not?” This question, meaning "Have you eaten?", corresponds exactly, in terms of syntax, to the Mandarin Chinese equivalent "吃了没有?” (chī lè méi yǒu):

Eat already or not?
没有?
verb perfective marker negation

In this sentence, “already” / “liao” corresponds to Chinese 了. In the Mandarin sentence, 了 marks completion and is thus a perfective aspect marker. “Already” / “liao” similarly functions as a perfective aspect marker in this sentence.

An ambiguous use of “already” / “liao”, however, can be seen here:

I eat already
pronoun verb marker

In both the Singlish and Chinese sentences, “already” / “liao” / 了 might mark completion, or might mark change of state. If “already” / “liao” / 了 are being used to mark completion, the sentences mean “I ate”; if they are being used to mark a change of state, then the sentences mean “I am about to eat” and “already” / “liao” / 了 therefore instead mark the inchoative aspect.

In Singlish, there are ways to disambiguate the meaning of the sentence, such as through the use of mood particles or interjections. It should be noted that in these situations, the primary function of these mood particles or interjections is not to disambiguate “already”, but rather that because they add additional context, they have the side effect of disambiguating “already” / “liao”. For example:

Eh I eat already can?
interjection pronoun verb marker question particle

(It is possible to contrive a scenario in which the above sentence means “I ate”, but it is a sufficiently marginal case that I will not consider it in this essay.)

In this example, the addition of the interjection “eh”, used to catch the attention of the listener, and the question particle “can?”, used in this case to ask for permission, provide sufficient context to determine that the speaker is asking for permission to begin eating, and therefore “already” / “liao” in this sentence serves as an inchoative aspect marker. A similar (but not syntactically identical) sentence in Mandarin Chinese, with the same meaning, would look like this:

可以 吗?
I can eat (marker) (question particle)
pronoun modal verb verb marker question particle

In the example above, 了 serves as an inchoative aspect marker as well.

Yeah I eat already
interjection pronoun verb marker

In the above sentence, the addition of the affirmative “yeah” suggests that the speaker is answering a question, possibly the question “Eat already or not?” or “Have you eaten?” and “already” here can only be a perfective aspect marker.

Another way to disambiguate the meaning of the sentence “I eat already” is to move “already” from sentence-final position to pre-verb position and to add “liao” in sentence-final position:

I already eat liao
已经
pronoun adverb verb marker

He already go home (liao)
已经
pronoun adverb verb object/adverb marker

(In the first example, “liao” is not optional, whereas in the second, “liao” is optional; in both Mandarin sentences, 了 is required. At the moment, I do not have an explanation for why “liao” is optional in some cases and required in others.)

In these examples, the syntactical correspondence with the Chinese sentence suggests that pre-verb “already” is not an aspect marker but is instead an adverb, corresponding to Mandarin 已经 (yǐ jīng). In this situation, the presence of adverbial “already” indicates that the action occurred in the past, and “liao” is therefore a perfective aspect marker. We can safely conclude that in Singlish, whenever “already” appears in pre-verb position, it is parsed as an adverb rather than as an aspect marker and the perfective aspect is implied, regardless of whether the verb is conjugated in the past tense, and if sentence-final “liao” is also present, it marks the perfective rather than inchoative aspect. Because pre-verb “already” corresponds to Mandarin 已经 rather than Mandarin 了, the Hokkien “liao” can never appear in pre-verb position. Additionally, “I already eat already” is understood but generally not accepted, due to the repetition of “already”. This is the only scenario in which “liao” cannot be replaced by “already”: when “already” appears elsewhere in the clause as an adverb.

Other verbs or adverbs can be used to disambiguate the function of “already” / “liao”. For example, “start” implies that “already” / “liao” marks the inchoative aspect:

I start eating already
开始
pronoun auxiliary verb non-finite verb form marker

Conversely, “finish” implies that “already” / “liao” marks the perfective aspect:

I finish eating already
pronoun auxiliary verb non-finite verb form marker

The above sentence has no syntactic equivalent in Mandarin. However, curiously, in Singlish, “finish” can be used as an adverb, analogous to Mandarin 完 (wán):

I eat finish already
pronoun verb adverb marker

Throughout this essay I have emphasized the correspondence of “already” / “liao” and Mandarin 了. However, there is one important difference between “already” / “liao” and 了: Mandarin allows the use of 了 in post-verb position, as in the following example:

已经
He already eat (marker) rice
pronoun adverb verb marker object
He has already eaten.

Singlish, however, never admits “already” / “liao” in post-verb position (unless that also happens to be the sentence-final position). Given that English does not admit “already” in post-verb position except after “be”, “have” (when used as an auxiliary verb) and modal verbs, I believe this is likely to be due to the influence of English syntax.

An additional note is necessary: in his thesis, Lim Heng Liang found there is no statistically significant difference between Chinese and Malay speakers in terms of which positions of “already” were acceptable for a given aspect, except for perfective sentence-initial “already”, which Malay speakers were more likely to accept than Chinese speakers. I would like to note that “already” is in fact acceptable in sentence-initial position in English, but it seems to me that any sentence beginning with “already” is likely to be parsed by a Singlish speaker as English rather than as Singlish, and its validity as a sentence will be analyzed as such. Furthermore, Malay does admit “sudah” (already) in sentence-initial position, but a deeper analysis of such a construction is beyond the scope of this essay.

To sum up, Singlish “already” / “liao” generally functions as either a perfective or inchoative aspect marker, indicating the completion of an action or a change of state, much like 了 in Mandarin. The ambiguous aspect can be clarified by the use of verbs or adverbs, such as “start”, “finish” or even “already” itself, that imply one of these two aspects. Unlike Mandarin 了, “already” / “liao” is acceptable only in sentence-final position, not in post-verb position, possibly as a result of English not accepting “already” in the post-verb position. Additionally, “already” is also acceptable in pre-verb position in Singlish, where it is parsed as an adverb, as it would be in English, rather than as an aspect marker. As such, the syntax of word “already” is a salient example of the multiple linguistic influences present in Singlish.


Works Cited

  • Lim Heng Liang. “Creole Homogeneity in Multiple Substrate Situations: Findings from Singapore English kena, already and until.” Undergraduate thesis, National University of Singapore, 2012. Print.
  • Ross, Claudia; Sheng Ma, Jing-heng. Modern Mandarin Chinese Grammar: A Practical Guide. New York: Routledge, 2006. Print.