This article didn’t turn out the way I originally planned it. The need to write this arose from comments I have gotten over the years, where readers criticized my way of labeling the members of a sentence – especially the subject. While I don’t need to justify my choices, I decided it would be nice to have an article to link to related to the topic.

My original goal was to give some kind of overview of what the term “subject” means in a language like Finnish, with some background information about the different approaches. I did meet that goal, but I got carried away and dived far too deep into the topic. I have gone through a lot of research papers and grammar books while writing this article. You can find the bibliography at the bottom of this page.

There are a lot of grammar terms mentioned in this article. I’ve tried to make sure I explain the terms I use and to give plenty of examples, but please understand that this article is mostly meant for those with some kind of linguistic background. On the other hand, those with such a background might notice that I have the tendency to use the inexact word “verb” rather than the more linguistical term “predicate”, and that I use the commonly understood term “sentence” most of the time, rather than the term “clause”. I do this in an attempt to make this article easier to read, but I’m aware that I’m fairly inconsistent.

In addition to this article, I will (soon!) be writing a second article which will be much more useful to the typical language learner. In that article, I will focus on what the current, most typical approach is to subjects.

Because the article is so long, I’m using a lot of boxes that hide part of the text until you click on them. I hope this helps you stay on track while reading this article. The issue with long articles like this is always how to display things as clearly as possible. Usually, I try to make each section similar in presentation, but I’ve been working on this article for so long that there is variation in the way I present the content. At this point I honestly can’t be bothered anymore to homogenize things.

Claim: “Every sentence must have a subject.”

One very simple example to refute this “rule” is something you’re certainly already familiar with. In Finnish, you can leave out certain pronouns to create subject-less sentences (e.g. “Minä itken.” could just be “Itken.”). This is called subject omission, which is typical for null-subject languages. In addition, there are certain sentence types in Finnish that just never have a subject in the first place (e.g. Sataa. On kylmä. “It’s raining. It’s cold.”). Finnish isn’t a “subjektivaltainen” (subject-prominent) language. Thus, less strict approaches towards the concept of a subject are commonplace.

Claim: “The subject is always located before the verb.”

Word order as a way to indicate the subject of a sentence is also under question because Finnish word order is quite free. The subject of many sentences can be freely moved to a less prototypical place. For example, the sentences “Ari lyö Juhoa.“, “Juhoa Ari lyö.” and “Juhoa lyö Ari.” all mean that Ari hits Juho, though the first version is the most prototypical. The subject Ari can be moved to different locations within the sentence because the partitive case form Juhoa shows us that Juho is the object.

In addition, word order has functions in Finnish that, in other languages, would be expressed with definite and indefinite articles (in English “a” and “the”). In Finnish, new information will in a neutral context generally be placed at the end of a sentence. This affects, for example, the meaning and information structure of existential sentences (more about this in section 6).

In addition, changing word order will affect, first, what cases you will be using in the sentence and, second, the required verb form (e.g. Metsässä on hiiriä vs. Hiiret ovat metsässä vs. Hiiriä on metsässä). This issue will be discussed in more detail in section 6.

Claim: “The subject is always inflected in the nominative case.”

Finnish has many cases, which add complexity when we try to pinpoint the subject of a sentence. As you will see in this article, not all subjects in Finnish appear in the nominative case (e.g. Poikia tulee. “Boys are coming.”). Likewise, there are plenty of situations where another element of the sentence appears in the nominative case rather than the subject (e.g. Talo rakennetaan. “The house is being built.”).

Thus, the idea that a subject – and only a subject – can have the nominative case ending doesn’t apply very well to Finnish grammar. This is not a novel concept at all. E. N. Setälä in his Suomen kielen lauseoppi (1884) already mentions partitive subjects in addition to nominative subjects in the context of partial subjects in existential sentences (see section 6).

Claim: “A subject always triggers subject-verb agreement.”

The term “subject-verb agreement” refers to the tendency for the verb of a sentence to be conjugated according to the person and number of the subject. This mechanism explains why the verb itkeä is conjugated in different forms here: “Minä itken, sinä itket, me itkemme.” (I cry, you cry, we cry).

Finnish has multiple sentence types where the verb is always conjugated in the third person singular. For these sentence types, the person or number of the subject doesn’t affect the verb conjugation at all. This is the case, for example, for existential sentences (e.g. Huoneessa on mies/miehiä. – see section 6) and possessive sentences (e.g. Minulla/Meillä on koira. – see section 7).

Many linguists over the decades (e.g. Setälä-Sadeniemi 1952, L. Hakulinen 1979) have specified that the plural nominative case subject can have subject-verb incongruence when the verb expresses the existence of something, or the coming into existence or cessation of existence. In addition, if we consider spoken Finnish a valid variant (e.g. Sivonen 2019), we have to take into account that the third person plural form is rarely used in sentences such as “Miehet juoksee∼juoksevat” (The men are running.).

Claim: “The subject expresses the agent of the predicate.”

While in a neutral prototypical sentence, the subject indeed expresses who or what does an action, it’s more complicated than that. Even Setälä (1884) already addresses that sentences can have a “logical” subject which is different from its grammatical subject. Logical (or psychological) subjects are related to how we perceive the meaning and focus of a sentence, while grammatical subjects are based purely on the syntax and morphology of the sentence. Setälä mentions possessive sentences (Isällä on kirja.), necessity sentences (Minun pitää mennä.) and feeling verbs (Minua janottaa.) in this regard.

Semantics are not my area of expertise (I find it a little bit pointless to ponder meanings in great detail). It is irrefutable, however, that the grammatical subject of a sentence often isn’t the agent of an action. Let’s have an example of thematic relation theory to prove that a subject’s role doesn’t always perfectly match up with the syntactic function the element fulfills.

While, often, the AGENT of a construction is the subject of the sentence, that is not an automatic one-to-one equivalence. The subject of a sentence can fulfill a variety of roles, such as agent, experiencer, beneficiary, theme or instrument. For example, for the sentence “Nainen teetti uuden mekon räätälillä.” (The woman had a new dress made at the tailor’s.), the subject is nainen, but the AGENT is räätälillä. The semantic role nainen fulfills is the one of AIHEUTTAJA. Likewise, the KOKIJA of a sentence can be the subject (Mies on juovuksissa. “The man is drunk.”), but it can also be an adverbial (Minusta tämä on hyvä idea. “I think this is a good idea”.).

Thus, I think it’s clear that roles don’t perfectly match up with the syntactic function an element of a sentence fulfills. Thus, not all subjects are agents, and not all agents are subjects.

When can we have a partitive subject?

The conversation regarding the subject of existential sentences didn’t start out very controversial: it was readily decided that koira was the subject. The focus, instead, was on defining if/when you can have a partitive subject (e.g. Huoneessa on maitoa. Ruokaa on pöydällä. Poikia juoksee pihalla.). This led to discussions aiming to define the parameters in which a partitive subject can be used, starting with the types of verbs and eventually the sentence types

In the beginning, linguists (e.g. Setälä 1884, Siro 1943) primarily discussed existential verbs (i.e. verbs that express that something exists somewhere, comes into being, ceases to be or changes into another state of being). Ikola (1954) added to the discussion by paying extra attention to word order in sentences with a partitive subject, concluding that partitive subjects only appear in existential sentences. The relation between word order and meaning was also widely discussed (e.g. Siro 1964a).

Do note that the term eksistentiaalilause (existential sentence) remained undefined in part of the discussion. This caused a lot of disagreement surrounding the definition and word order in sentences that express something existing, coming into existence or ceasing to exist. Even the name of this sentence type was up for discussion for a while: Ikola (1954) named it “eksistentiaalilause“, while Penttilä (1963) used the term “eksistenssilause“. These two linguists in particular had a long debate going between them (circa 1954-1957) regarding existential sentences.

As I mentioned in the previous section, early on, linguists readily accepted that koira is the subject. However, as the decades passed, it became clear that not everyone found this so obvious.

Maybe koira is the object?
Lauri Hakulinen (1926) presented the sentence “Suomessa on kylmät talvet.” as proof that the T-plural is the object. The sentence doesn’t have subject-verb agreement, so kylmät talvet can’t be a subject. He considered the T-plural to be the accusative case rather than the nominative case. Hakulinen amended his opinion later in his career (1979), when he mentions the term “partitive subject” in existential sentences.

Maybe koira is the subject?
One issue concerning existential sentences and objects relates to personal pronouns (e.g. minä, sinä). In regular object sentences, such as “Minä muistan teidät“, the object te is inflected in the accusative case (teidät rather than te). If we compare this object sentence to existential sentences, we notice that it is impossible to say “Turussa on/asuu teidät“. Some linguists (e.g. Osmo Ikola 1954) see this as proof that existential sentences can’t have an object. Thus, an alternative approach is to consider koira the subject of the sentence “Huoneessa on koira“.

Or is koira the object after all?
There was a lot of going back and forth between subject and object interpretations. Kalevi Wiik (1974) gives a great overview of the different possible interpretations of what clause member koira could be in “Huoneessa on koira“. Wiik himself is in favor of interpreting koira as the object of the sentence. His main argument is based on looking past the “surface subject” and considering the “deep subject” (following Noam Chomsky’s transformational grammar approach). According to Wiik, “Huoneessa on koira” is to be understood to mean the same as “Huone sisältää koiran” (The room contains a dog), where koira is a clear object.

Perhaps koira is the complement?
Certain linguists have not been in favor of creating new types of subjects with separate rules. They feel it is futile to dig for a subject in atypical sentences. In anything but a prototypical sentence, it makes more sense to use other terms besides the subject. Linguists such as Jorma Toivainen (1986) consider koira to be the complement (predikatiivi) in an existential sentence. Thus, there is no subject in these sentences – which he didn’t consider to be a problem. After all, Finnish has a lot of sentence types that don’t include a subject at all.

Existential sentences have an existential subject

The currently most common approach (e.g. Leino 1989, Sorjanen 1999, Löflund et al. 2010) is to consider koira the subject. Since the release of VISK in 2004, the term e-subjekti (i.e. eksistentiaalisubjekti “existential subject”, eksistentiaalilauseen subjekti “existential sentence’s subject”) has become an easy way to specifically refer to the subject of an existential sentence. This separate subject type has become commonplace because of the conclusion that existential sentences just don’t have a typical subject; they require a different approach. The term e-subjekti isn’t VISK’s invention; it’s already mentioned in Hakulinen and Karlsson’s Nykysuomen lauseoppi (1979).

Helasvuo and Huumo (2015) reject the idea that existential sentences must contain a subject. They argue for a syntactic category called the e-NP (existential noun phrase), separate from the concept of the subject. Their theory is based on discourse analysis. They reason that the existential NP (koira) isn’t the subject due to the way it is used in the discourse.

They give an e-NP the following characteristics:

• It’s a noun phrase (NP), so a noun, a pronoun, or the combination of an adjective and a noun.
• It normally is the semantic first argument of the verb (i.e. the most important element in the sentence that helps complete the meaning of the verb, according to valency theory).
• It does not trigger subject-verb agreement.
• It appears in either the nominative or the partitive case.
• It’s usually positioned after the verb.
• Its referent is usually new information, which is mentioned but usually not tracked further in the discourse.

Note how other linguists (e.g. VISK) choose to attribute all these characteristics to what they call the e-subject. Only the very last element of my description above is specifically the result of Helasvuo’s and Huumo’s approach of the e-NP through discourse analysis. They draw attention to how the subject of a standard sentence is typically a recurring element in the discourse, and usually an active doer. In an existential sentence, on the other hand, the referent is often just used to characterize the locative phrase (huoneessa). For existential sentences, the new referent is usually (91 %) non-human, and often (71 %) doesn’t become a recurring element in the discourse. These percentages come from Marja-Liisa Helasvuo’s article “Ollako vai eikö olla – eksistentiaalilauseen subjektin kohtalonkysymys” (1996). It was a pretty interesting read, and it’s in English, so if you can get access to this book, do have a look!

1. No mention of the term “subject” at all
Some textbooks for second language learners (like you and me) don’t mention the term “subject” at all. They just present existential sentences as a construction and focus on the cases used, for example:

2. Subject underlined: Huoneessa on koira.
Those textbooks that do include the term subject, consider koira the subject of “Huoneessa on koira“, and readily mention the partitive as a possible case for the subject to appear in.

Early on, there were two camps when it came to analyzing possessive sentences: some found the case ending (nominative) the most important indicator, others focused on subject-verb agreement. Linguists in favor of the former concluded that koira is the subject, while those who were in favor of the latter considered koira the object.

Koira is the subject
Setälä (1884) had a nominative-centered approach for this sentence type. In his sentence “Isällä on kirja” the subject is kirja because it appears in the nominative case. This is a clear example of how, early on, linguists considered the case used to be the main indicator that marks a word as the subject in a sentence. However, even Setälä already mentioned that the possessive sentence “Isällä on kirja” has a logical subject (i.e. isä) which is different from the sentence’s grammatical subject (i.e. kirja).

Koira is the object
Lauri Hakulinen (1926) considered subject-verb agreement proof that koira is the object rather than the subject. He uses the example “Pojalla on vaaleat hiukset.“, where the verb is conjugated in the singular third person regardless of the T-plural form used. This is typical for objects: they typically don’t have any effect on the form of the verb that’s used. Thus, Hakulinen marked the locative noun phrase (pojalla) as the subject, and the clause member in the nominative as the object (vaaleat hiukset). He also mentions the sentence “Minulla on sinut.“, where we have the accusative case form sinut. This case is normally used for objects, which also speaks in favor of Hakulinen’s interpretation of the possessed entity as the object of the sentence.

Koira is the subject
Ikola responds to Hakulinen in a Virittäjä article (1954), where he makes it clear he considers the possessed entity to be the subject. To reach his conclusion, he compares several existential and possessive sentences as well as some necessive existential sentences (Talossa on isäntä. Talossa pitää olla isäntä. Minulla on saappaat. Säkissä on saappaat.). He states that these sentences all have an identical structure, which means they must all be interpreted in the same way. Existential sentences have their subject at the end of the sentence, so he concludes that the same is true for possessive sentences. This is in contrast with Hakulinen, who considers possessive sentences and existential sentences to be two separate sentence types with their own rules. As to why we have the accusative case in “Minulla on sinut“, Ikola states there’s no clear reason, and seems to think this is a new development in the Finnish language.

In interpretations where koira is considered the subject, meillä is seen as a habitive adverbial (habitiiviadverbiaali).

The linguist Vilkkuna (1996) gives a nice overview of everything that’s typical and atypical for both subject options (meillä and koira). She adds that if we consider meillä the subject, it’s an obliikvisubjekti (oblique subject, which in Latin grammar has meant any case other than the nominative and vocative, and in Finnish usually means one of the location cases, such as the adessive form meillä).

Many linguists (e.g. Löflund et al. 2010) compare possessive sentences and existential sentences to one another, seeing how they have many things in common (e.g. an adverbial at the beginning of the sentence (meillä/huoneessa), olla-verb, no subject-verb agreement). Nowadays, it’s considered logical to group both sentence types under one umbrella term because of their similarities. Thus, koira is the subject and meillä an adverbial.

Grammar sources like VISK (§ 923) choose to use the term e-subjekti in possessive sentences as well as in existential sentences. The term e-subjekti is short for eksistentiaalisubjekti “existential subject”, or eksistentiaalilauseen subjekti “existential sentence’s subject”. VISK makes the distinction, however, between “Minulla on auto.” and “Minulla on kuuma.” While the former is a possessive sentence, they consider the latter an ambient clause (see section 13). This classification makes sense from a semantical point of view: you don’t own anything; you’re just feeling the heat. This also helps explain the lack of the partitive case in negative sentences (e.g. Minulla on kuuma. / Minulla ei ole kuuma.). Note that VISK only considers “minulla on” sentences with an adjective ambient clauses. This means “Minulla on nälkä/jano/kiire.” (hungry/thirsty/in a hurry) are still considered possessive sentences, even though they also are exempt from the rule to have partitive in negative possessive sentences.

1. No mention of the term “subject” at all
Most textbooks for second language learners (like you and me) don’t mention the term “subject” at all. They just present this sentence type as, for example: “Missä? + olla + auto”, “Missä? + ei ole + autoa”. This approach avoids the problematic nature of the subject completely.

2. Subject underlined: Meillä on koira.
Those textbooks that do include the term subject all mark the final word of the possessive sentence as the subject. Thus, the choice is purely grammatical rather than taking semantics into account:

“Lähteä” as the subject

Seeing the infinitive of a verb as the subject of a sentence could be considered the “traditional” approach. Setälä (1901) already mentioned infinitives as subjects as an option in addition to nouns, adjectives and subordinate clauses. He mentioned the verbs pitää, täytyy, tarvitsee, tulee and sopii as examples of impersonal (persoonaton) verbs which have an infinitive subject. Many other linguists over the decades (e.g. Siro 1964, Konttinen 1978, Leino 1989) have considered this a valid approach.

Traditionally, the genitive “minun” in the sentence “Minun täytyy lähteä.” was considered a “dative adverbial” (e.g. Setälä 1960, Leino 1989). A dative adverbial is a part of speech which expresses to whom or which something is happening. The term “dative-genitive” (datiivinen genetiivi, datiivigenetiivi) has also been used (e.g. Penttilä 1963, Ikola 1964). According to these theories, the form minun is separate from a standard genitive because it carries a dative meaning (the obligation of leaving falls to minun).

Sorjanen (1999) remarks that there are two interpretations in use for necessity sentences. In the main text, he marks the verb’s infinitive as the subject (Lasten pitää päästä hampurilaisbaariin.). In a footnote, however, he mentions that for pedagogical purposes, it makes sense to speak of a genitive subject for this sentence type, as the Ministry of Education and Culture has done in 1993. It’s important to keep in mind who a grammar interpretation is meant for.

Noam Chomsky’s transformational grammar led to the idea that we have several “layers” in sentences. The idea is that, in the sentence “Minun täytyy lähteä.“, there is a “hidden second sentence”: “Minä lähden.“. The minä-subject of that hidden sentence is completely prototypical, which according to these linguists (Ikola 1964, Siro 1964b) meant that for the necessity sentence, the genitive word is the subjektipersoona of the infinitive. Rephrased, according to the transformational grammar theory, the genitive is the infinitive’s subject (i.e. the subject of lähteä), rather than the subject of the predicate täytyy.

This leads linguists (e.g. Hakulinen–Karlsson 1979) to naming minun the “grammatical surface subject” (kieliopillinen pintasubjekti). The infinitive can either be considered to serve as an attribute of the main verb (i.e. [täytyy [lähteä]] with brackets showing the dependency), or as part of the verb chain (i.e. [täytyy lähteä] as a whole).

Vilkuna (2000) states that it’s possible to interpret the genitive minun in this sentence type in two ways. Firstly, we can consider the genitive the infinitive’s subject. Secondly, we could simply say the genitive is the necessity verb’s subject.

Most commonly these days, linguists (VISK § 921, Ikola 2001, Löflund et al. 2010) state that this sentence type has a genitive subject. The genitive subject (genetiivisubjekti) is only used in necessity sentences. Its features are that it is inflected in the genitive case, and generally appears in front of the verb that expresses necessity. When the subject refers to one person or thing, it will be inflected in the singular genitive (e.g. Kokin täytyy pilkkoa sipulit.). It can also be inflected in the plural genitive form (e.g. Kokkien täytyy pilkkoa sipulit.).

Necessity sentences can also be used without the genitive form person. For example, “Täytyy lähteä.” is a complete sentence, which depending on the context either means that the speaker (minun) must go, or can refer in a generic kind of sense to multiple people or people in general. Another added dimension is brought on by “existential necessity sentences”: “Täällä pitäisi olla paperia.” (There should be paper here), which are a combination of two sentence types: something must/should be in a location. I will be exploring this mixed sentence type in a future article.

1. No mention of the term “subject” at all
Textbooks for second language learners (like you and me) don’t always mention the term “subject” at all. They just present this sentence type with examples. This helps avoid the matter of identifying the subject completely.

• Suomi sujuvaksi 1: “Genetiivi + TÄYTYY/EI TARVITSE + 1. infinitiivi”
• Suomen mestari 2: “G + täytyy/ei tarvitse + infinitiivi”
• Ahaa! 1: “Kenen? + täytyy + verbin perusmuoto”
• Alku 1: “Kenen? + täytyy + verbin perusmuoto”
• From start to Finnish: “Kenen -n (genitive) + täytyy/pitää + 2nd verb, basic form”

2. Subject underlined: Minun täytyy lähteä.
Those textbooks that do include the term “subject” unanimously mention that these sentences have a genitive subject.

• Alku 2: “Subjekti eli verbin tekijä on genetiivissä.”
• A Grammar Book of Finnish: “The subject of a necessive clause is in the genitive case.”
• Suomen kielioppia ulkomaalaisille: “Nesessiivilauseen subjekti (tekijä) on genetiivissä.”
• Finnish: A Comprehensive Grammar: “The genitive is the case of the SUBJECT with some verbs of necessity or obligation (täyty-y ‘must’, on pakko ‘have to’, etc.), and some verbs with a modal meaning (e.g. kannatta-a ‘be worth (doing)’ (sth.), sopi-i ‘may’, onnistu-u ‘succeed’).”
• Suomi sujuvaksi 2: “Tekijä (subjekti) genetiivi + verbi yksikön 3. persoonassa + toinen verbi infinitiivi / TAVA-partisiippi + (objekti/adverbiaali) + (adverbiaali/objekti)”
• Sun suomi: “The subject of these sentences is in the genitive, no verb gets a personal ending.”
• Suomen kielen tikapuut. Alkeistaso 1: “Modaaliverbit täytyä ja pitää ilmaisevat välttämättömyyttä. Ne esiintyvät aina yksikön 3. persoonassa ja niiden subjekti on genetiivissä.”

Sentences such as “Minua pelottaa.” (I’m afraid) typically have three elements: a verb that expresses a feeling or emotion, an experiencer (kokija) in the partitive case, and the cause of the emotion (ärsyke, aiheuttaja) in the nominative case. However, this is a sentence type where some verbs can appear without any other element. For example, the sentence “Pelottaa.” (I’m scared.) works fine on its own. When used without an explicit person, the experiencer of the feeling is usually implicitly referring to the speaker themselves. In addition, some verbs simply can’t have a causative agent. For example, “Minua nukuttaa.” will never have a nominative case element added that functions as an agent causing the feeling.

For the simple sentence “Minua pelottaa.” (I’m afraid.), we could interpret minua both as the subject or the object. Translating to English definitely makes it look like a subject, though it’s not the doe-er of the sentence but rather the experiencer. Then again, pelottaa is a causative verb, so it is also possible to consider minua the object to which this cause is applied.

This sentence type becomes even more interesting when we add an agent to it: “Minua pelottaa sota.“. Now, we can see that, semantically, sota is the agent that brings about the emotion, while minua is the experiencer of the emotion. With the addition of sota, the subject interpretation for minua becomes less plausible, because sota has more prototypical elements that are normally associated with the subject.

The word sota can both be located at the end and at the beginning of the sentence: “Minua pelottaa sota.” vs. “Sota pelottaa minua.“. A prototypical sentence of this type will have the experiencer (in my example minua) at the beginning of the sentence. I find the English translations interesting because the subject of the English sentences “I’m scared of war.” and “War scares me.” changes depending on word order, while Finnish free word order doesn’t have an effect on what we might consider to be the subject.

This sentence type can have the following possible schema (muotti):

1. Minua pelottaa. “I’m scared.”
2. Minua pelottaa sota. “I’m scared of war.”
3. Sota pelottaa minua. “War scares me.”
4. Sota pelottaa. “War scares (people).”
5. Pelottaa. “(I’m) scared. ~ This is scary.”
6. Minua pelottaa lähteä sotaan. “I’m afraid of going to war.”
7. Minua pelottaa, kun ihmiset puhuvat sodasta. “I’m scared when people talk about war.”
8. Minua pelottaa se, että sota saattaa levitä Suomeen. “I’m scared that the war could spread to Finland.”

Many linguistic descriptions (e.g. Huhmarniemi 2019a, Siro 1978) focus on the roles of the elements of the sentence. In our sentence “Minua pelottaa sota.“, the NP (noun phrase) at the beginning of the sentence is the experiencer, while the NP at the end of the sentence is stimulus or cause of the emotion or feeling. Much of the discussion is centered around which element of the sentence occupies a higher position in the argument structure using the methods of dependency grammar and valency theory. Thus, we have multiple linguists (e.g. Hakulinen-Karlsson 1979) who analyze these sentences as some version of: “partitive case noun (experiencer) + verb + nominative case noun (stimulus/cause)”. They focus on which of the two elements is the verb’s primary argument.

Most linguists (e.g. Setälä 1901, Hakulinen A. 1972, L. Hakulinen 2000, VISK § 467) agree that sota is the subject of the sentence “Minua pelottaa sota“. Thus, the partitive case experiencer of the sentence is the object.

It is important to point out that this sentence type doesn’t have an obligatory subject, if we decide to name sota the subject in “Minua pelottaa sota.“. It’s, for example, very common to see the verb ärsyttää used like “Minua ärsyttää.“, without any further element added to the sentence. Setälä (1960) refers to these as “temporarily (nominative) subject-less verbs”.

This is also the first sentence type on this page where a verb infinitive can be construed as the subject of the sentence: Minua pelottaa lähteä. (I’m afraid to leave.). The approach to name an infinitive as a subject has been accepted by many linguists over the decades (e.g. Setälä 1884, Siro 1978, Konttinen 1978, Ikola 1991).

VISK (§ 503) also argues just that: lähteä is the subject because the infinitive fills the same semantic function as the NP subject, and can thus be replaced by a noun. There are of course many things that set infinitives apart from regular subjects. For one, we can’t inflect an infinitive in the nominative or partitive case nor in the plural to test for subject-verb agreement. In addition, we can’t move the infinitive to the beginning of the sentence, which is the prototypical location of the subject.

In addition to an infinitive, this sentence type can also get a subordinate clause as a subject. Most commonly, this is a subordinate että-sentence (e.g. Minua pelottaa, että sota voisi tulla myös Suomeen. “I’m afraid the war could come to Finland as well.”). Some linguists (e.g. Huhmarniemi 2019b) note that we could add the pronoun se (essentially a dummy pronoun here) to fill the role of the subject: Minua pelottaa se, että sota tulisi myös Suomeen. It’s also possible for the subordinate clause to be an indirect question (e.g. Minua huolestuttaa, tuleeko sota myös Suomeen. “I’m worried about whether the war will come to Finland as well.”).

1. No mention of the term “subject” at all
Many textbooks for second language learners (like you and me) don’t present this sentence type at all. When presented, they are often approached just through examples. Usually, the term “subject” isn’t mentioned at all.

• Suomen mestari 3: “P (Ketä?) + verbi (hän-persoona) + nominatiivi/infinitiivi/sivulause”. The construction isn’t given a name, the title just says “Minua ärsyttää“
• Suomi sujuvaksi 1: “Tunnekausatiivilause: lauseen verbi on esimerkiksi väsyttää, nukuttaa, janottaa […]. Ketä väsyttää? Minua väsyttää. Minua ei väsytä. Sinua nukuttaa. Sinua ei nukuta.” There is no mention of sentences where there could be a noun in the nominative case added at the end of the sentence.
• Työelämän suomea 2: “Kiinnostaa on kokijaverbi, kuten minua harmittaa/janottaa. Kokijan täytyy olla partitiivissa.”

2. Subject underlined: Minua pelottaa sota.
Those textbooks that do include the term “subject”, consider the word at the end of the sentence which causes the emotion/feeling to be the subject. The fact that this sentence type can also have a subordinate clause or an infinitive which expresses the cause of the emotion/feeling is generally not mentioned.

• Suomi sujuvaksi 2: “Kokija (objekti) Partitiivi + Verbi + (Mikä?) (subjekti).”
  → Proposed subject underlined: “Minua pelottaa sota.”
• A Grammar Book of Finnish: “Clauses without a subject: […] 4. The clause has a verb that expresses a feeling or physical state. NB! The partitive (päätä, minua) in these clauses is grammatically the object. These sentences can also have a subject: Pekan vitsi (subject) nauratti minua (object). Pekka’s joke made me laugh.”
  → Proposed subject underlined: “Minua pelottaa sota.”
• Finnish: A Comprehensive Grammar: “Clauses without a subject […] 5. Causative verbs expressing feelings ‘(Minua [*mua]) harmi>ttaa’. ‘Harmittaa, että ulkona sataa.’ The että ‘that’ clause is here the subject of harmittaa.”
  → Proposed subject underlined: “Minua pelottaa, että sota tulisi myös Suomeen.”

Some linguists (e.g. Hakulinen–Karlsson 1979, Vilkuna 2000, Sakuma 2007) have used the idea of dislocation to analyze quantifying sentences. Dislocation means that two elements that are located apart from one another really form one entity. When analyzing quantifying sentences using this approach, these linguists explore what types of problems arise when we put the dislocated element back together. They show that changing word order changes the sentence type and also has an effect on the cases used.

For example, if we attempt to turn “Poikia oli kolme.” into an existential sentence by changing the word order, we get “Oli kolme poikaa.“, with the singular partitive rather than the plural. We can’t turn it into a regular sentence with a T-plural subject either, because “Pojat olivat kolme.” is an invalid sentence. All this is to say that quantifying sentences are clearly their own sentence type rather than an example of how word order in Finnish is free.

Siro (1964a) remarked that these are existential sentences where the place has been left out. For example, “Väkeä oli paljon.” is the shortened version of “Väkeä oli siellä paljon.“. From this, he draws the conclusion that väkeä is the subject of the sentence, and paljon an adverbial. This is the approach that’s currently widely accepted.

VISK (§ 902) includes quantifying sentences under the umbrella term of existential sentences, thus naming the partitive case word at the beginning of the sentence an e-subjekti (existential subject). The sentence “Oppilaita oli neljä.” has the following things in common with existential sentences: the “subject” (oppilaita) is inflected in the partitive case, and the verb is either olla or another existential verb. The following thing sets it apart from typical existential sentences: word order. The subject is located before the verb rather than after. VISK accepts the partitive case as a possible case ending for the e-subject, so it doesn’t mark this as atypical for this sentence type.

Not all quantifying sentences have a subject. All the examples I’ve mentioned so far had a partitive case subject (if we choose to interpret it like that) because their verb is intransitive (e.g. olla and tulla). If the verb is transitive (e.g. kutsua), the partitive case noun is considered the object of the sentence. Compare, for example, “Syitä oli monta.” and “Syitä keksittiin monta.” (There were a lot of reasons. vs. A lot of reasons were thought up).

I must say I lost interest in the conversation between linguists about this language type. There were at least three approaches: “Oppilaita oli neljä.” could be “subject + verb + complement”, “subject + verb + adverbial” or “adverbial + verb+ subject”. I agree with Hakulinen–Karlsson (1979), whose (translated) sentence states: “It’s difficult if not impossible to unequivocally solve, what functions the constituents of quantifier sentences have in the surface structure”.

Most textbooks for second language learners (like you and me) simply don’t mention this sentence type. For those that do, one avoids the term subject at all, and the other considers the quantity to be the subject:

• Suomi sujuvaksi 2: “Meitä on monta (kvanttorilause): Partitiivi + Verbi yksikön 3. persoonassa + Kuinka paljon?”
  → Proposed subject underlined: Vastauksia tuli neljä.
• Finnish: A Comprehensive Grammar: “Quantified clauses with initial partitive nominal, and a quantity
  expression as subject: ‘Auto-j-a [*autoi] oli kolme.'”
  → Proposed subject underlined: Vastauksia tuli neljä.

Through linguistic analysis, the proposed options for interpreting change-and-result sentences are either “adverbial (-stA) + verb + subject” (e.g. Siro 1964a) or “adverbial (-stA) + verb + complement” (e.g. Penttilä 1963, Löflund et al. 2010). Either of these two interpretations seems possible, especially when we take into account that the final member of a change-and-result sentence can be both a noun (Lapsesta tuli lääkäri.) and an adjective (Lapsesta tuli oikullinen.).

Compare:

Siro (1964a) considers the final word in the sentence the subject. In order to account for the noun/adjective options, he proposes to use the term subjekti for nouns, and the term “predikatiivinen subjekti” (complement subject) if the subject is an adjective. This interpretation helps explain why a negative change-and-result sentence requires the partitive case. We can compare “Lapsesta ei tullut lääkäriä.” and “Kaupasta ei tullut lääkäriä.” and see the parallel.

VISK (§ 904) considers the result of the change the complement (predikatiivi), both when it’s a noun and an adjective. The elative-case member at the beginning of the sentence is considered an adverbial. Thus, this sentence type does not have a subject at all.

A problem with interpreting the last member of this sentence type as a complement is that it doesn’t exactly behave like one. Namely, the case used does change to the partitive in a negative sentence. Compare, for example, the sentences “Lapsesta tulee lääkäri. Lapsesta ei tule lääkäriä.” to the typical complement sentences “Lapsi on lääkäri. Lapsi ei ole lääkäri.“

While linguists compare change-and-result sentences to both complement sentences and existential sentences, they (e.g. Hakulinen–Karlsson 1979, Siro 1964a, VISK § 891) consider it important to consider this sentence type to be its own type rather than a subcategory of one of the other types.

1. No mention of the term “subject” at all
It’s quite rare for beginner textbooks for second language learners (like you and me) to mention this sentence type at all. A couple of books present this sentence type with a few examples, give a basic idea of the structure, but then don’t name the sentence type at all:

• Suomen mestari 3: “sta/stä + tulee” and “sta/stä + ei tule + P”. This book doesn’t name the sentence type: the title of the grammar page just says “Minusta tulee…”
• Suomi sujuvaksi 1: “Lausetyyppi: Minusta tulee lääkäri: -STA + tulla-verbi yksikön 3. persoonassa + mikä? mitä? millainen? millaista?”

2. Subject underlined: Lapsesta tuli aikuinen.
Those textbooks that do include the term “subject”, all consider the final word of the sentence as the subject:

• Suomen kielioppia ulkomaalaisille: “Kenestä? Mistä? + Tulla-verbi (yksikön 3. persoona) + Mikä? Mitä (subjekti)”.
  → Proposed subject underlined: Lapsesta tuli aikuinen.
• A Grammar Book of Finnish: “A subject in the partitive case is only possible in the following clause types: existential (“there is / are”), possessive and change + result clauses (elative construction).”
  → Proposed subject underlined: Liisasta ei tullut lääkäriä.
• Suomi sujuvaksi 2: “Kenestä? Mistä? Elatiivi + Tulla-verbi (yks. 3. pers.) + Mikä? Mitä? Subjekti“.
  → Proposed subject underlined: Lapsesta tuli aikuinen.
• Finnish: A Comprehensive Grammar: Result clauses with adverbial and subject: ‘Pojasta kasvoi mies.”
  → Proposed subject underlined: Lapsesta tuli aikuinen.

The consensus among linguists is that infinitives and subordinate clauses can be considered the subject of an “on hauskaa” sentence. As one of the first, Setälä (1901) embraced this idea. This goes to show that the nominative-first approach of international linguistics immediately got multiple caveats from the very beginning of Finnish syntactic research.

Most sources (e.g. Hakulinen–Karlsson 1979, Vilkuna 2000) point out that infinitives and clauses have very little in common with prototypical subjects (i.e. no verb-subject agreement, no noun in the nominative case, not placed at the beginning of the sentence). However, infinitives and subordinate clauses in “on hauskaa” sentences are used as primary arguments the same way noun phrase subjects are.

Some linguists (e.g. Airila 1947, Sorjanen 1999) use the base term subjekti for any type of subject, including infinitive or subordinate clause subjects. Other sources (e.g. Leino 1989, Vilkuna 2000) prefer to keep these less prototypical subjects separately. They use the terms infinitiivisubjekti (infinitive subject) and lausesubjekti (sentence subject) to distinguish them from the regular subjekti.

VISK does not have a separate “On hauskaa” sentence type. They consider this sentence construction to be a type of ambient clause (see section 13). Following this approach, sentences describing the weather (e.g. On sateista. “It’s rainy.”), the state of a place (e.g. On vaarallista. “It’s dangerous.”) and the way people experience a situation (e.g. Oli mukava tavata. “It was nice to meet.”) are all combined. VISK considers the infinitive the subject and the adjective the predikatiivi (complement) of the sentence (VISK § 504).

1. No mention of the term “subject” at all
“On hauskaa” sentences are mentioned in some textbooks for second language learners. When this sentence type is included, you usually don’t find the term “subject” in the description:

• Suomen mestari 1: Has three example sentences underneath the heading “On hauskaa…”. Elsewhere in the same book: “on/ei ole + adj. P + inf.”
• Suomi sujuvaksi 1: “Olla-verbi + millaista? mikä? + tehdä mitä? / että”
• Suomi sujuvaksi 2: “On hauskaa matkustaa (on hauskaa-lause): (Adverbiaali Missä? Milloin Millä?) + Olla-verbi yksikön 3. persoonassa + Predikatiivi Millaista? Millainen? Mikä? + A-infinitiivi/että-lause” + elsewhere in the same book: “Olla-verbi yksikön 3. persoonassa + predikatiivi + A-infinitiivi/että-lause”
• Sun suomi: “This structure describes what it is like to do something. Verb “on” works together with an adjective and a verb”

2. Subject underlined: On hauskaa uida.
I found two textbooks that include the term “subject”:

• Finnish: A Comprehensive Grammar: “‘Oli mukava juoda kahvia.’ Infinitive phrase (InfP) as subject. ‘Oli mukavaa, että Meeri tuli.’ Clause as subject.” + “A clause is a complex sentence also if it has an infinitive phrase embedded in a central syntactic function, e.g. as subject: On hauska uida ‘It is fun to swim’.”
  → Labels the infinitive or subordinate clause as the subject (On hauskaa uida. Oli mukavaa, että Meeri tuli.)
• A Grammar Book of Finnish:
  Contains two approaches: “We use the infinitive or a subordinate clause (starting with conjunctions such as että, jos, kun) after constructions of the on hauskaa type. The infinitive and the subordinate clause are the grammatical subject of this type.” Elsewhere in the same book: “The agent in clauses of the type On hauskaa can also be regarded as a subject in the genitive.”
  → Labels either 1) the infinitive or subordinate clause as the subject (On hauskaa lähteä lomalle. Onpa kivaa, että saitte asunnon.) or 2) the optional genitive element as the subject (Minun on vaikea oppia suomea.)

This sentence type has multiple possible structures, which range from very common to pretty rare. These different structures have been used as a way to analyze this sentence type and the function of the elements in it. Kolehmainen (2010) takes a very interesting deep dive into the syntactic and semantic roles of weather verbs. She starts out by presenting both “Pilvi satoi vettä.” (The cloud rained.) and “Pilvestä satoi vettä.” (It rained from the cloud.). Note that these are both very unusual ways to talk about rain, but they’re not impossible. The former sentence places pilvi in the position of the subject, while in the latter pilvestä is an adverbial.

Next, here’s a compilation of the different weather sentence structures linguists have mentioned:

• Sataa. “It is raining/snowing.”
  – Typical usage of the verb sataa.
• Sataa lunta. “It is snowing.” (literally: raining snow)
  – Sataa + lumi (partitive “some snow”)
  – Typical usage with a partitive case element that could be analyzed as an existential subject (e-subjekti) or an object.
• Pilvestä sataa lunta. “From the cloud it is snowing.”
  – Pilvi (elative “from the cloud”) + sataa + lumi (partitive “some snow”)
  – Less typical, but possible. Strengthens the interpretation that lunta is an existential subject because it looks more like a prototypical existential sentence: it has a locative element at the beginning of the sentence.
• Pilvi sataa lunta. “The cloud is snowing.”
  – Pilvi (nominative “the cloud”) + sataa + lumi (partitive “some snow”)
  – Even less typical, but possible. Pilvi should be interpreted as a prototypical subject here: it’s located before the verb, inflected in the nominative case and semantically the agent of the action. Thus, lunta must be the object. We can’t have two subjects or two objects in one sentence. Making the sentence plural strengthens this interpretation: “Pilvet satavat lunta.” (subject-verb agreement).
• Kattoa korjattiin sen verran, ettei vesi sada sisään.“The roof was fixed to the extent that water no longer rained inside.”
  – Vesi (nominative “the water”) + ei sada (“doesn’t rain”) + sisään (illative case “to the inside”)
  – This sentence puts vesi in a situation which is prototypical for the subject: located before the verb, inflected in the nominative case and semantically the agent of the action.
• Satoi ensilumen. “It snowed the first snow.”
  – Sataa + ensilumi (accu-gen -n “the first snow”)
  – Unusual, but possible. The genitive (accusative) -n in this sentence marks ensilumen as the object of the sentence.

I agree with the conclusion Kolehmainen (2010) draws: meteorological sentences such as “Sataa.” can’t easily be considered prototypical subject-less sentences. Rain and snow can, depending on the speaker’s intent, be considered the agent of the action and, thus, behave like a prototypical subject. In a sentence like “Sataa lunta.“, however, lunta can’t be fully interpreted as either the subject or the object, as it carries neither the full meaning of the agent or the patient.

VISK (§ 457) also concludes that these verbs are not completely subject-less, mentioning that it’s possible to add a partitive subject (Ulkona pyrytti lunta. “Outside it snowed heavily.”) as well as a basic subject (Ensilumi satoi Lappiin. “The first snow came down in Lapland.”). VISK also states that weather verbs can be used with an object (Satoi ensimmäisen lumen. “It snowed the first snow.”).

We could interpret these sentences as complement sentences (predikatiivilause), where the subject has been left out. Alternatively, we could consider these sentences a type of existential sentence (eksistentiaalilause), where the adverbial has been omitted (e.g. Huoneessa on kuuma. “It’s hot in the room.”). An adverbial typically expresses a place, time or manner in which something happens.

Adjectives (e.g. kuuma) are more likely to be complements, which makes the first interpretation sound slightly more natural for kuuma. Nouns (e.g. kesä) are typically found in both complement sentences and existential sentences. However, the vital semantical feature of a tilalause is that it describes a sensation or ambient atmosphere. A complement sentence with a noun doesn’t do that. Rather, it identifies what something is (compare “Poika on pitkä.” and “Poika on opiskelija.“).

On the other hand, marking kesä in “Silloin oli kesä.” as the predikatiivi (complement) helps explain why making the sentence negative doesn’t make kesä partitive. The negative sentence “Silloin ei ollut kesä.” will have kesä in the nominative case. If this sentence was interpreted as a typical existential sentence, the existential subject would be inflected in the partitive case in this context (e.g. Huoneessa oli koira. Huoneessa eli ollut koiraa.).

As already mentioned, many linguists just don’t give a detailed description of this sentence type. Hakulinen and Karlsson (1979) group “Minulla on kuuma/nälkä.” under the same type as “Minua ärsyttää.” (see section 9) based on their meaning.

VISK (§ 900) combines “On kesä.” and “On kuuma.” by interpreting both the noun kesä and the adjective kuuma as complements. It is interesting how VISK uses the term predikatiivi (complement), but doesn’t consider “On kuuma.” a complement sentence. Instead, they consider this sentence to belong under the wider umbrella term of existential sentences (VISK § 891). Normally, VISK considers existential sentences to have an e-subject, so using the term predikatiivi in this case means they’re mixing two sentence types.

The approach Helasvuo and Huumo (2010) present is to abandon the terms subject, object and complement altogether for atypical sentence types (see more in section 6). For sentences like “On kuuma/kesä.“, we could just refrain from trying to compare it with the prototypical features of the subject, object and complement, and instead use the term e-NP (existential noun phrase).

Some linguists (Penttilä 1963, VISK § 906) bring forward that the genitive case can be used to add an experiencer to the sentence: “Lapsen on jano/kylmä.“. Nowadays, using the genitive case is less common, and considered more dialectical

Some ambient clauses can have the person experiencing the sensation at the beginning of the sentence in the adessive case (-llA): in addition to saying “On kuuma.” (It’s hot.), we can also say “Minulla on kuuma.” (I feel hot.). For some linguists, this changes its classification as a sentence type. According to VISK, adding an adverbial that expresses ownership (habitiiviadverbiaali) rather than a genitive turns these into possessive sentences.

Penttilä (1963) considers “Minulla on kuuma.” a possessive sentence just like “Minulla on auto.“. This is the same approach textbooks for second language learners use. For second language learners, it is a useful simplification at the start of your studies. However, it means these textbooks have to add an additional rule: the exceptional nominative/partitive variation in negative possessive sentences. A regular possessive sentence will have the partitive case (Minulla ei ole autoa.), while “Minulla ei ole kuuma/kylmä/jano/nälkä.” will have the nominative case (the basic form).

VISK makes things complicated by considering sentences with a noun (i.e. “Minulla on nälkä/kiire.“) possessive sentence (§ 895), and sentences with an adjective (i.e. “Minulla on kuuma/kylmä.”) ambient clauses ( § 900). While I understand that they are different in meaning, the fact that none of these will have the partitive case ending in a negative sentence unites them. Due to the different sentence type, VISK also analyzes the function of the final word in these sentences differently: nälkä and kiire are considered e-subjects in VISK, while kuuma and kylmä are considered complements.

1. No mention of the term “subject” at all
Most textbooks for second language learners (like you and me) don’t mention this sentence type by name at all. The weather is something that’s covered in most beginner course books, so meteorological sentences are always listed in that regard, but they are not named as a sentence type nor analyzed at all.

As for “On kuuma.” sentences, the “Suomi sujuvaksi” series does mention the sentence type by name, but doesn’t use the term “subject”:

2. Expressly state “no subject”
Some textbooks do list “On kuuma.” and “Sataa.” as a sentence type, placing them under the header of subject-less sentences. They don’t give the sentence type any name.

The term ilmiölause isn’t used in any textbook I looked at. Fred Karlsson does list the sentence “Syttyi sota.” under “existential clauses with adverbial and subject” in his book Finnish: A Comprehensive Grammar, without analyzing the sentence any further.

Features of the e-subject
I will be publishing an article dedicated specifically to the e-subject, but for now, here’s a list of its typical features according to VISK. I’m adding an example of both existential and possessive sentences. VISK considers possessive sentences to be a subtype of existential sentences.

• The cases used for the e-subject are the nominative and the partitive case (as well as their plural forms).
  • Huoneessa on koira. Minulla on koira.
  • Huoneessa on äidin silmälasit. Minulla on silmälasit.
  • Jääkaapissa on maitoa. Minulla on rahaa.
  • Jääkaapissa on marjoja. Minulla on ystäviä.
• In negative sentences, the e-subject will appear in the partitive case.
  • Huoneessa ei ole koiraa. Minulla ei ole koiraa.
  • Huoneessa ei ole äidin silmälaseja. Minulla ei ole silmälaseja.
• The e-subject doesn’t have subject-verb agreement. Rather, it will be conjugated in the third person singular.
• The e-subject will be located after the verb.