Is English a pro-drop language?

[dihydrogen monoxide]

English is a pro-drop language, but something about it puzzles me. Let's take this sentence:
a) I ate an ice-cream today.

According to pro-drop parameter, English language speaker could say something like
b) Ate an ice-cream today.

The word order of this sentence can be changed to
c) Today I ate an ice-cream

This is where the puzzle begins. If I have a sentence a), then I can drop the personal pronoun, but if I have a sentence c) I cannot drop the personal pronoun.

Why can't you drop a pronoun in sentence c), but you can in sentence a)?

 

[bibax]

English is considered a non-pro-drop language.

 

[Meat]

English is not pro-drop.

You might hear your sentence "b" in very casual speech, or in songs or something, but for the most part, you need to include the "I".

 

[dihydrogen monoxide]

Whether English is a pro-drop is debatable. We can say it's non pro-drop with pro-drop tendencies, ie behaves like one.

 

[CapnPrep]

Whether English is a pro-drop is debatable.

Uh, not really… As bibax and Meat pointed out, English cannot be considered pro-drop under any useful definition of the term. This does not mean that null subjects absolutely never ever occur in English. The precise conditions under which they can occur have been studied by many linguists. You can find some references to get you started by searching for "diary drop".

 

[Scholiast]

Good day, everyone

In the light of CapnPrep's observation...:

English cannot be considered pro-drop under any useful definition of the term

...I am moved to wonder whether there can be any "useful definition" of the term "pro-drop" at all, at least as regards the languages known to me - that is, apart from my native English, Latin, Greek, some French and Spanish and nearly fluent German. I also have a smattering of Russian and Turkish.

Vulgar - no value-judgment here, just a reference to the quotidian vernacular - English, French, German, Spanish and Russian speakers regularly "drop" otiose pronouns in speech where the formally written language would demand more intact and exact syntax. This phenomenon ranges from the compilation of shopping-lists to the composition of personal diaries to poetry and other "high" literature, where sometimes a writer takes, for artistic reasons, liberties with the requirements of formal grammar.

Of course I stand open to correction here, but I suspect that the "pro-drop" term is an invention by a scholiaster to make sound scientific what is in fact a common feature of the differences merely between spoken and written language in many parts of the globe.

 

[LilianaB]

Polish, for example is a language where the use of personal pronouns as subjects would be considered bad style in many cases, or even most of the cases, probably. Lithuanian is also a pro-drop language. Many of the sentences would sound unnatural with the pronouns. You cannot say Ja jade do pracy, in literary Polish, but jade do pracy, only. Perhaps in very few situations this would be allowed, as an answer to the question who being one of them. This applies to the first person singular and plural. It varies with other pronouns.

 

[Roy776]

a) I ate an ice-cream today.

According to pro-drop parameter, English language speaker could say something like
b) Ate an ice-cream today.

The word order of this sentence can be changed to
c) Today I ate an ice-cream

This is where the puzzle begins. If I have a sentence a), then I can drop the personal pronoun, but if I have a sentence c) I cannot drop the personal pronoun.

Why can't you drop a pronoun in sentence c), but you can in sentence a)?

First of all: It's not like you could write sentence b) without any previous context, either. You need to know beforehand "who" ate the ice-cream. At least my feeling is telling me so. I could imagine that in colloquial speech.
- "What did you do?" - "Ate 'n ice-cream."
- "What did he do?" - "Ate 'n ice-cream."
The sentence alone doesn't work well, although I immediately considered it first person singular.

And in regards to sentence c):
This just shows that English is NOT a pro-drop language, as pro-drop languages are able to drop the pronoun in almost every tense and especially every sentence structure.

Whether English is a pro-drop is debatable.

It is not.

Good day, everyone

...I am moved to wonder whether there can be any "useful definition" of the term "pro-drop" at all, at least as regards the languages known to me - that is, apart from my native English, Latin, Greek, some French and Spanish and nearly fluent German. I also have a smattering of Russian and Turkish.

Let's take a look at some examples of both phrases in some languages I know (corrections appreciated).

German: a) Ich habe ein Eis gegessen. b) Heute habe ich ein Eis gegessen.
Here you could drop the pronoun to change the sentence. "Hab' heute ein Eis gegessen." This is vulgar and works only because Hab represents the first-person singular conjugation.
Spanish: a) Comí un helado. b) Hoy comí un helado.
You could say "Yo comí un helado" but it is unnecessary, as Comí already represents the first-person singular indefinido. Note: In Spanish it is in no way colloquial or vulgar to drop the pronouns.
Polish: a) Zjadłem lody. b) Dzisiaj zjadłem lody.
Just the same as Spanish. You can drop the pronouns and it's neither colloquial nor vulgar. Same counts for Czech, but as it's almost the same, Polish should suffice.

Furthermore, in pro-drop languages, the use of pronouns sometimes even makes sentences seem incorrect or badly formulated and they are more or less just for emphasis.

- He preguntado, ¿quién comió el helado?" - "Yo lo comí." (Yo = I)
- "Zapytałem, kto zjadł lody?" - "Ja zjadłem go." (Ja = I)


So, what to consider a pro-drop language, and what not? Easy, if you ask me. In a pro-drop language, said phenomenon is neither vulgar nor colloquial, but an accepted and always intelligible standart.

 

[LilianaB]

In Polish the main drop of the pronoun in the above mentioned sentence would be Ja as the subject. You drop the subject which is absolutely necessary if one wants to sound natural. It is not a colloquialism, but quite to the contrary. Ja ide, would be considered colloquial, but most likely improper use. Ja zjadlem lody could be only an answer to Who ate the icecream. So, I agree with Roy.

 

[Scholiast]

Dear all

"Hab' heute ein Eis gegessen." This is vulgar and works only because Hab represents the first-person singular conjugation.
Spanish: a) Comí un helado. b) Hoy comí un helado.
You could say "Yo comí un helado" but it is unnecessary, as Comí already represents the first-person singular indefinido. Note: In Spanish it is in no way colloquial or vulgar to drop the pronouns.
Polish: a) Zjadłem lody. b) Dzisiaj zjadłem lody.
Just the same as Spanish. You can drop the pronouns and it's neither colloquial nor vulgar. Same counts for Czech, but as it's almost the same, Polish should suffice.

This was precisely what I was referring to. "Suppression", perhaps better simply "Omission", of the pronouns, is common throughout the I-E languages, at least when they are spoken.

German, Spanish and the Slavonic languages differ however from English in this much, that the inflected form of the verb usually indicates the grammatical person anyway.

This leaves me, however, so klug wie zuvor, as to the scientific definition, if there is such a thing, of "pro-drop" as a technical term in Linguistics or Philology.

Best to all

 

[Roy776]

This leaves me, however, so klug wie zuvor, as to the scientific definition, if there is such a thing, of "pro-drop" as a technical term in Linguistics or Philology.

Wikipedia provides us with this definition:

A pro-drop language is a language in which certain classes of pronouns may be omitted when they are in some sense pragmatically inferable (the precise conditions vary from language to language, and can be quite intricate). The phenomenon of "pronoun-dropping" is also commonly referred to in linguistics as zero or null anaphora.

 

[Dan2]

Of course I stand open to correction here, but I suspect that the "pro-drop" term is an invention by a scholiaster to make sound scientific what is in fact a common feature of the differences merely between spoken and written language in many parts of the globe.

I have no specialized knowledge of "pro-drop", but if you look at the familiar languages English, French, and German on the one hand, and Spanish and Italian on the other, there seems to be a huge difference in the appropriateness of dropping the subject pronoun in most circumstances, whether in speech or writing, justifying proposing a categorical distinction.

And the ability to pro-drop in a language appears to be only partially related to whether the verb carries the subject information:
- John never even finished high school.
- Yes, but I've noticed that ___ speaks very well. (in speech or writing)

 

[Roy776]

I have no specialized knowledge of "pro-drop", but if you look at the familiar languages English, French, and German on the one hand, and Spanish and Italian on the other, there seems to be a huge difference in the appropriateness of dropping the subject pronoun in most circumstances, whether in speech or writing, justifying proposing a categorical distinction.

And the ability to pro-drop in a language appears to be only partially related to whether the verb carries the subject information:
- John never even finished high school.
- Yes, but I've noticed that ___ speaks very well. (in speech or writing)

Of course it's only partially related to it, just take a quick look at japanese. There are almost no pronouns in use, although the verb does never contain the subject. But some languages simply allow the omission of pronouns and others don't. German sounds always almost awkward when pronouns are dropped. You can't say long sentences without pronouns, simply impossible. "Was hast (du) gemacht?" (lit. What hast (thou) done?), okay, but longer sentences are just impossible and awkward, so I'd say that the consideration of a pro-drop language depends highly on the language's own standarts. As already mentioned, while such sentences are perfectly fine in Spanish, Polish and Czech, they are not in German, Swedish and others.

 

[berndf]

...I suspect that the "pro-drop" term is an invention by a scholiaster to make sound scientific what is in fact a common feature of the differences merely between spoken and written language in many parts of the globe.

No, it is a technical term in Generative Grammar where it is a useful term and a meaningful concept. ... If you are a true believer in the universal grammar hypothesis.

 

[CapnPrep]

If you are a true believer in the universal grammar hypothesis.

It is true that non-generativist researchers may choose to avoid the term "pro-drop" because of its close historical association with the pro-drop parameter. However, it is still a convenient term for referring to the linguistic phenomenon of omitting contextually salient unstressed pronominal arguments, which is widespread throughout the world's languages (as Scholiast rightly points out), but which is subject to conditions that can be very different from one language to another (as the other posters rightly point out). One can use the term "pro-drop" without necessarily believing in a universal parameter or a categorical distinction. It has the (questionable) advantage of sounding scientific, but it was not invented with that sole purpose in mind.

 

[berndf]

However, it is still a convenient term for referring to the linguistic phenomenon of omitting contextually salient unstressed pronominal arguments...

No objection. I am not a Chomsky fan and I still use the term.

My point, in reply to Scholiast, was that in the context of GG it is much more than a convenience or a fancy way of talking but a theoretical concept that requires a specific term.

 

[Alxmrphi]

In linguistic typology it's still alive and well, and not connected to any Generative theory at all.
English is definitely not considered pro-drop in the linguistics community.

 

[terredepomme]

Interesting thread. I don't know if the subjects being omitted in French can be considered a similar phenomenon because they are never "completely" dropped but retain their last consonants or are phonetically "merged" with the following verb. E.g.

Je sais pas. - Chais pas. (I don't know. - Dunno.)
Vous êtes fous? - Z'êtes fous? (Are you crazy? - You crazy?)

It is also interesting to remark that in Korean and Japanese, even objects can be removed, and they very often are if the context allows it. For example

사랑해Saranghae. - (I) love (you).

To say literally "I love you"(나는 너를 사랑해) would be really awckward and unadvisable.
Also note that there is almost no notion of person-based verb conjugation in Korean. The sentence does not grammatically imply the first-person singular, all we have is the context.
So the same sentence could mean "(you/he/she/they/etc) love (me/him/her/them/etc)" based on the context.

 

[Ghabi]

It is also interesting to remark that in Korean and Japanese, even objects can be removed, and they very often are if the context allows it.

Hi Terre. I think Japanese and Korean (and Chinese and Thai and many other Eastern-Asian languages for that matter) are considered topic-prominent, and may not compare well with the European languages mentioned above, which are subject-prominent (pro-drop or not).

 

[gburtonio]

I've often thought about this. I think an argument can definitely be made for 'pro-dropness' operating along a cline. In terms of the principles and parameters theory I believe it's supposed to be a binary property, i.e. languages either are or aren't pro-drop. I'd say the following part of the Wikipedia definition is debatable:

'certain classes of pronouns may be omitted when they are in some sense pragmatically inferable (the precise conditions vary from language to language, and can be quite intricate)'

I say debatable because the 'precise conditions' don't just vary from language to language, but within languages. So in English we have the 'diary drop' phenomenon, as well as the tendency to drop subjects in spoken English. All because the subject can easily be inferred. As speakers we are generally going to want to express ourselves in as efficient a way as possible (with exceptions, e.g. literature, rhetorical speech etc. where how you say something may become more important than what you say), so dropping any pragmatically redundant part of an utterance seems logical.

You can look at these phenomena as being exceptions to the rule – so English is a pro-drop language except that subjects are dropped in certain circumstances. But I think you could also say (and this would be my view) that a language can drop a subject any time it's clear from the context. In languages that have verb conjugations that clearly mark the subject, this will always be possible. In English, the context alone will mean it is sometimes possible.

I think the sentence:

*Today ate an ice-cream

is only unacceptable because the context now becomes ambiguous – is 'today' an adverb or the subject? There is some ambiguity there (admittedly small), so the subject is needed.

 

[Roy776]

I think the sentence:

*Today ate an ice-cream

is only unacceptable because the context now becomes ambiguous – is 'today' an adverb or the subject? There is some ambiguity there (admittedly small), so the subject is needed.

I wouldn't be so sure about that, as the same phenomenon also occurs in German, although the verb "haben" in its conjugated form already represents the first-person-singular.

"Hab' heute ein Eis gegessen." "Ich hab' heute ein Eis gegessen."
"Heute hab' ein Eis gegessen." "Heute hab' ich ein Eis gegessen."

The first sentence can even be used without any prior questioning or discussion of the topic. The false sentence represents the "phenomenon" and the two sentences on the right are correct High German.

And to show the difference in conjugation:
1.p.s. habe
2.p.s. hast
3.p.s. hat
1.p.p. haben
2.p.p. habt
3.p.p. haben

As you can see, one would still know who or what the subject is, but it's still not acceptable, even in the worst vulgar speech.

 

[Dan2]

so English is a pro-drop language except that subjects are dropped in certain circumstances.

non-pro-drop except...?

But I think you could also say (and this would be my view) that a language can drop a subject any time it's clear from the context. In languages that have verb conjugations that clearly mark the subject, this will always be possible. In English, the context alone will mean it is sometimes possible.

This make sense, but I don't think it's correct. It's exactly this view that I argued against in post #12. In the dialog,
- John never even finished high school.
- Yes, but I've noticed that ___ speaks very well.
note that
a) the verb is unambiguously marked for 3rd person singular
b) the context makes it 100% clear who the subject is
c) other languages have no problem with subject-drop here ("che parla bene")
And yet in English, even in the most informal speech, we simply must include "he".

 

[Alxmrphi]

Great point Dan.
There are even languages which have completely full verbal conjugations and still require subjects (like Icelandic), like in Italian you'd easily say "che parla bene", you just couldn't say "En ég tek eftir að ____ talar mjög vel". You'd need "hann" (he). It allows more droppings than English does but you can't say that it's anywhere near like a typical pro-drop language, so even then it's not "always" about the ability for the verb to carry the subject information, like in Dan's example the context makes it so clear as to who the person being talked about is, but it just would never be said. That is obviously the reason why the languages that do it can do it, because subject information is conveyed via the verb, but just because they can doesn't mean they do.

 

[Roy776]

I hope this isn't beyond the scope of this topic, but wouldn't it be useful to compile a list of languages that are pro-drop languages, ordered by their language families? This could show wether certain language families have a tendency of dropping pronouns or not.

 

[gburtonio]

OK, it's true that I drastically oversimplified things. Embedded clauses clearly create extra restrictions. I still think it operates along a cline, however. Clearly it's not just context that allows a subject to be dropped, but to me it seems more appealing to think of English as being a certain point along a cline of 'prodropness', and that there are certain restrictions (contextual and grammatical) on when subjects can be dropped. That to me just feels instinctively closer to the truth than saying it's a non-pro drop language (binary value) which for some reason frequently drops subjects. I suppose fundamentally I don't see there's anything intrinsically different linguistically in saying 'Can't hear you.' on the phone and an Italian saying 'Non ti sento.'

I also think that imperatives work the same way. Why don't we use subjects in imperatives? Because it's obvious that the subject is the person we're talking to.

 

[Roy776]

I suppose fundamentally I don't see there's anything intrinsically different linguistically in saying 'Can't hear you.' on the phone and an Italian saying 'Non ti sento.'

But at least I see the difference in the languages' rules! Italian allows for pro-dropping in its literary standart, while English does not. Of course, language rules could be rewritten, but who of us (especially us native speakers) would actually accept such radical changes?

 

[gburtonio]

But at least I see the difference in language's rules! Italian allows for pro-dropping in its literary standart, while English does not. Of course, language rules could be rewritten, but who of us (especially us native speakers) would actually accept such radical changes?

Hmm, I think when you start talking about what's 'allowed' in a literary context, then you're not really talking about the same thing any more. And literature in any case would allow for subject dropping in the context of dialogues. Can't diaries also be a form of literature?

What would be fascinating (although of course impossible) would be to find some illiterate native English speakers, teach them to write (purely the skills to transcribe speech) and see if when transcribing they always included subject pronouns. If they did, that would suggest that the pronouns are always there but simply aren't always articulated (I think this would be the Chomskyan explanation). If they didn't, it would suggest that the fact that in written English they're almost always present (apart from in dialogues, diary entries etc.) is simply literary convention that doesn't reflect the realities of the language.

 

[Alxmrphi]

If they did, that would suggest that the pronouns are always there but simply aren't always articulated (I think this would be the Chomskyan explanation).

That's the Chomskyian explanation anyway with actual pro-drop languages (it's called 'PRO/pro' i.e. here).
There is a point to making a distinction between pro-drop and non pro-drop languages, and English simply doesn't fit anywhere near the category of a pro-drop language, so there's no point in classifying it based on some 'diary-dropping' (I forget what name it's been given, something to do with a diary though) feature of language.

Aha, here's an article on Subject Dropping in Informal English, you might find interesting.
I see it's 'diary-drop' I was trying to remember, almost got it!

 

[gburtonio]

That's the Chomskyian explanation anyway with actual pro-drop languages (it's called 'PRO/pro' i.e. here).
There is a point to making a distinction between pro-drop and non pro-drop languages, and English simply doesn't fit anywhere near the category of a pro-drop language, so there's no point in classifying it based on some 'diary-dropping' (I forget what name it's been given, something to do with a diary though) feature of language.

Aha, here's an article on Subject Dropping in Informal English, you might find interesting.
I see it's 'diary-drop' I was trying to remember, almost got it!

I don't disagree with what you say, but the idea of pro-drop (and maybe the ideas have been significantly updated since the original version of the principles and parameters theory – I'm admittedly not up to date) was that it is a binary parameter, which is set either one way or another. I'm just saying that to me it instinctively makes more sense to conceive of languages not as either being one or the other (even if you do start adding sub categories like 'little pro'), but to see subject dropping as a phenomenon that can happen in many languages but in different circumstances and with different restrictions.

 

[Alxmrphi]

I'm just saying that to me it instinctively makes more sense to conceive of languages not as either being one or the other (even if you do start adding sub categories like 'little pro'), but to see subject dropping as a phenomenon that can happen in many languages but in different circumstances and with different restrictions.

And people do, but for Chomsky it had to be binary and his explanation is limited to his followers and people who have tried to go further.
P&P has died a death in Linguistics, but some of the terminology has been revamped into Minimalism and earlier branchings off the original G&B framework, and pro-drop and the analysis of it has remained around to describe languages which systematically don't require pronouns. The word "Viene?" is a sentence in Italian, it truly does not need a subject if the people understand who it is about, if you say "Coming?" in English, you've got no verbal information about who it is and you're restricted to take the immediate assumption it means the person being spoken to, which is a very different case from the Italian version.

Many people view it like you do, and they probably call it "Not needing to use pronouns all the time in languages". It seems an exception is being taken to the naming of a characteristic, and judging it in the realm of the real world when it never was intended to be thought of in that way. There certainly is a case for a cline-like scale that languages sit on, but "pro-drop" as a strictly binary feature belongs to a world and a context that nobody is in anymore, so it's applicability can't be judged by just looking around and trying to typologically explain languages. It's like saying "The Verb-Attraction parameter is useless because it doesn't help to explain how to learn the language", which is a true statement, but it's taking the term and applying it and judging it to a completely different context, it makes sense in the realm it's supposed to be in.

So I think once we remember now that it was a term used in initial-studies into P&P and is a useful classification of linguistic features around the world (hence its continued use in language typology), but taking special cases and applying the term to it isn't really useful as it was never designed to be judged like that, and be exception-free. Parameters were more often about generalisations than exceptionless rules, the exceptions were always taken to be explained by an external feature.

But I do get your point on the issue

 

[merquiades]

As Roy suggested in really colloquial English the first person "I" can often be dropped if the context is clear. This normally happens only in dialogues or maybe monologues too (like letters).
- Hope you have a nice day.
- Love you so much honey.
- Wish it would rain today.
- Went to the bank today but it was closed.
It certainly cannot be said that pronouns in general can be dropped routinely at will like in Spanish or Italian. Stylistically it's better to include the "I" always.

 

[jmx]

I'd like to mention that in Spanish, and probably in all Romance languages save French, subject pronouns not only can be 'dropped', they must be dropped in some contexts, in order to sound natural. There are many posts in the Spanish Forums of WR where you can immediately spot that the poster is not a native of Spanish, just by their excessive use of subject pronouns.

 

[Alxmrphi]

I'd like to mention that in Spanish, and probably in all Romance languages save French, subject pronouns not only can be 'dropped', they must be dropped in some contexts, in order to sound natural. There are many posts in the Spanish Forums of WR where you can immediately spot that the poster is not a native of Spanish, just by their excessive use of subject pronouns.

Yeah it's sort of the same in Italian, a non-native speaker transfers pronoun-usage from English and thinks it sounds normal, but in Italian it's actually putting a big emphasis on the statement, which can be considered rude in a lot of contexts, stressing for another person to do something or what they themselves want. What sort of examples did you have in mind where it'd be considered incorrect in Spanish?

 

[jmx]

Yeah it's sort of the same in Italian, a non-native speaker transfers pronoun-usage from English and thinks it sounds normal, but in Italian it's actually putting a big emphasis on the statement, which can be considered rude in a lot of contexts, stressing for another person to do something or what they themselves want. What sort of examples did you have in mind where it'd be considered incorrect in Spanish?

Well, I wouldn't say "be considered incorrect" but rather "sound unnatural".

Let's put a simple example. You go to a restaurant alone and the waiter asks about what you would like for dinner, and then you answer "yo quiero una sopa" (literally, "I want a soup"). The waiter would probably look at you with an expression of surprise, one that he wouldn't show if you had said "quiero una sopa". The explanation is that the first sentence, "yo quiero una sopa", is better translated in that context as "as for me, I want a soup", which of course could be a natural thing to say if you were at the table with other people, but not if you were alone, as in the example.

 

[merquiades]

I'd like to mention that in Spanish, and probably in all Romance languages save French, subject pronouns not only can be 'dropped', they must be dropped in some contexts, in order to sound natural. There are many posts in the Spanish Forums of WR where you can immediately spot that the poster is not a native of Spanish, just by their excessive use of subject pronouns.

You're right of course. Subject pronouns are almost a misnomer in Spanish and Italian. I almost want to call them disjunctive pronouns. The subject is contained in the verb ending, the pronoun just gives emphasis. I'd contrast it with French where subject pronouns are so present they might seem redundant at times. In "Je vais au cinéma", the "je" cannot disappear even if "vais" is only used for the first person verb. They lack total emphasis, are always present matter of fact, and often elided in speech "j'vais", so much so that the disjunctive pronouns often need to be added as well: "Moi, je vais au cinéma." or "Je vais au cinéma moi." That's more or less the equivalent of "Yo voy al cine" or "Voy yo al cine". French is the opposite of a pro-drop language.
I can't think of a way to express this subtlety in English though, unless we just physically give emphasis to the word. I'm going to the theatre vs. II am going to the theatre.

Now that I think of it, it's strange that subject pronouns are required in German, even though spoken verb endings give all the needed info. Ich spreche, du sprichst, er spricht, wir sprechen etc.

 

[terredepomme]

"Moi, je vais au cinéma." or "Je vais au cinéma moi." That's more or less the equivalent of "Yo voy al cine" or "Voy yo al cine". French is the opposite of a pro-drop language.I can't think of a way to express this subtlety in English though, unless we just physically give emphasis to the word. I'm going to the theatre vs. II am going to the theatre.

I think this is a phonetical difference between French and English. French doesn't have relaxed vowels like in English so it's difficult to emphasize specific words in a sentence, but in English, you could say II am going to the theatre, I am GOING to the theatre, I am going TO the theatre, I am going to the THEATRE etc, to mark the abovementioned subtlety. In French it would depend less on this phonetical emphasis but more on syntax such as moi, je vais au cinéma. Or in other cases such as l'état, c'est moi whereas in English one would say I AM the state.

 

[muhahaa]

About the requirement of subject pronouns even though there is conjugation: http://en.wikipedia.org/wiki/Standard_Average_European

 

[Salanikola]

English subject pronouns are always dropped in the imperative mood/modality (except as vocative interjections). "You eat an ice-cream." is a statement, but "Eat an ice-cream." is a command (or advise). "You! Eat an ice-cream." and "Eat an ice-cream, you! " are also commands (odd ones), but the "you!" is a separate interjection not the subject of the verb. Traditional grammar calls this the "implied subject", or "understood you". The imperative is not always used as a command per se, it can also be used for requests and indirect speech acts that aren't questions (ie. "Let me tell you ...") and can be combined with the adverb please for (usually) polite requests, but the imperative (in which the verb is always a "naked infinitive") is the only case where pronoun dropping is standard in English. ("naked infinitives" are just present tense verbs, except that be is the "naked infinitive" of to be)

Also, I would agree that "Ate an ice-cream." would sometimes be used by some speakers to answer a question (and maybe occasionally in other situations like casual diaries), while "Today ate an ice-cream." would not (in any dialect I know). I think "Ate an ice-cream today." is just as good as "Ate an ice-cream." except that it is strange as an answer to a question. I think the main reason it is strange is that the context is usually given by the question or already known before the question is asked.
I think that (maybe) when English speakers drop pronouns in this way, they think of it more like saying a fragment of a sentence then as dropping individual words. That's probably why "Today ate an ice-cream." is so totally unnatural to English speakers, while "Ate an ice-cream." is occasionally used. The first one feels like you made the sentence shorter (by dropping "I") and then made it longer again by adding a word ("today") in a place that it would normally never be (right before the verb).
Keep in mind mind that many speakers rarely or never drop pronouns (outside the imperative), even when answering questions or writing diaries.

That's just my intuition and experience as a native speaker.

EDIT: Actually subject pronouns are not always dropped the imperative. The archaic construction <naked infinitive> + thou/ye/you + <rest of predicate> is common in old works like the King James Bible and in the cry "Hear ye! Hear ye! Hear Ye!" used to call court rooms to order. It is occasionally used by people trying to be facetiously formal/archaic and in fiction writing trying to sound archaic. A similar construction: <verb> <obj. pronoun> <rest of predicate>, may or may not be imperitive, and is a dialectical/archaic alternative to using reflexive pronouns that can be used in any tense. Examples of this second (unrelated) construction include: "Get thee to a nunnery." (imper., subj. = implied thou, thee = dir. obj.) and "I'm'a get me a new car." (indic., subj. = I, me = ind. obj.) ("I'm'a" = I am going to = alternative future tense).

 

[Salanikola]

Actually, I agree with what I think is one thing the article that Alxmrphi shared was saying. Oftentimes when English speakers "drop" pronouns it's more like we're just knowingly failing to pronounce them. I think we usually imagine the pronounce on some level before we say the sentence, even if we know full well we're not saying it.

Q: "What is that?"
A: "... Don't know."

Out of the blue.
"[I'm] going to the cinema."

If I were the answerer, I would probably have had no pretense in my mind of actually saying the word "I", but I also probably would have thought it as I said the sentence. The article seems to be saying that it's similar to contractions in English, though for me personally, it feels rather different. (I don't think "I am" when I say "I'm") This might be dialectical or the result of education though.

There is another thing I noticed though:

Q: "What are you doing?"
A: "Eating ice-cream." (most common) or "I'm eating ice-cream." (almost as common), or "M'eating ice-cream." but never "Am eating ice-cream" (for most speakers).

This might be because the whole of "eating ice-cream" is essentially an adjective (participial phrase?) (and looks exactly like a noun (gerund) phrase), and adjectives and noun phrases absolutely can occur alone as answers to questions in spoken English, if they're replacing the "question word" ("what" in this case I think, even though "doing" also is needed in the question). Many speakers will also use verbs without subjects (but possibly their other arguments) as answers to questions:

Q: "What did you do when you woke up?"
A: "Brush my teeth." or "Brushed my teeth." or "I brushed my teeth."
Q: "What will you do in the interview?"
A: "Just be myself." or "Be myself.", or "I'll just be myself." or "I'll just be myself." or maybe "I will just be myself."

Even people who do this (which I think is most English speakers) probably wouldn't say "Am eating ice-cream", the full word "am" only exists in the more formal/enunciated version of this sentence ("I am eating ice-cream.") while the word "I'm" is used in the informal unenunciated form. You will sometimes hear things like "M'eating ice-cream.", with either a syllabic or a non-syllabic [m]. (usually non-syllabic if before a vowel)

That last parentheses meant "(It is usually non-syllabic if it is before a vowel.)", and was an example of dropping both subject and verb, similar to the "Eating ice-cream." example. In writing, I believe this type of clause is supposed to be separated by a comma from the rest of a sentence, rather than being considered a separate sentence, but it confuses me when it's a parenthetical so I often write it that way.