Motivation
One aspect of Esperanto grammar I am still struggling with, is verbs. I am by no means the only one. I do remember almost 3 years ago, I went back from the Universala kongreso in Turin after taking my B1 exam and being confused about the semantics of igi/iĝi suffixes. During the multi-hour ride, my friend and I built up a basic understanding along common explanation practices. About one year ago, I mentioned that I want to sum it up in a blog post and another Esperantist told me to share it, because they get confused about it as well. I do believe the accusative case and transitiveness are the two most common issues Esperantists struggle with. Unlike the former, I found the latter cannot be explained easily and to some extent they are intertwined.
Let’s dive into the topic of “how do Esperanto verbs operate?”
Verbs
Fundamentally, verbs in a language describe actions or relations between entities. “I love you” yields a (at least unidirectional) love between ‘I’ and ‘you’. “I became an Aikido trainer” yields that ‘I’ attained a certain property (or relation) ‘being an Aikido trainer’.
Linguistic concepts
If we look at other languages, there are various associated concepts. I am going to use languages I am familiar with here.
-
In German, one can say “Ich backe einen Kuchen” (en. “I bake a cake” or “I am baking a cake”). In my elementary school, one of the biggest deals was understanding that ‘Ich’ is a subject and ‘einen Kuchen’ is an object. The concept of subject/object is fundamental, but my favorite linguist taught me that this concept is language-specific and easily misunderstood. But we are going to stick to it here. In this German sentence, the nominative element is unambiguously ‘Ich’ (accusative would be ‘mich’) and the accusative element is ‘einen Kuchen’ (nominative would be ‘ein Kuchen’). So we just define that the subject is the element in nominative case and the object is in the accusative case. And in Esperanto we can use the same terminology, since every accusative element is distinguishable from the nominative element by the -n suffix.
-
In German, I can turn around the sentence and hide the subject: “Ein Kuchen wird gebacken” (en. “A cake is baked” or “A cake is being baked”). Here, we have no accusative. Recognize that the object in the previous example is the subject in this example. This form is what is called passive voice. There is no similar concept like this in Esperanto.
-
The previous examples showed an interesting concept: there is an acting entity (‘Ich’) and some entity acted upon (‘Kuchen’). Linguists call it agent and patient. This is semantic property and does not change by the example sentence structures. In both cases, the cake is the patient and the agent is I (even if it is only implicit in the second example). The sentence “I moved the pen to the table edge” has more than two entities. ‘I’ am acting on ‘the pen’ and the final entity ‘table edge’ might need another thematic role besides agent or patient. More semantic roles need to be defined. Manfred Krifka lists more roles for German like an instrument or a locator. Whereas the agent/patient model is basic and easy to explain, advanced models require explanations beyond what we try to achieve here.
-
Christer O. Kiselman wrote an elaborated text “Transitivaj kaj netransitivaj verboj en Esperanto” covering the topic of this blog post. He uses the word valento (en. valence). The verb is enriched by one, two, or more entities. By sentence structure and their convention, you know how those entity relate to each other. Whereas the agent/patient/… model assigns semantics to each element, the valence-model only counts and does not necessarily assign anything. So to understand Esperanto verbs better, I suggest to study its thematic roles as well.
-
I just claimed “the verb is enriched by one, two or more entities”. Does Esperanto have verbs without entities? Does Esperanto have complete sentences only with one verb? Yes. Esperanto defines that weather-related sentences may skip the subject. “Pluvas” means “It is raining” and skips the generic ‘It’, we find an English.
-
When breaking it down like Kiselman, one could claim that verbs have a valence of one or two, but this is missing crucial information. We neither specify the semantics (e.g. agent/patient) nor is it a complete model (i.e. covers verbs with zero entities). So I oppose explanations of verbs which emphasize whether the verb is transitive (has valence 2) or intransitive (has valence 0 or 1).
To conclude:
-
Esperanto verbs require zero, one, two, or more entities.
-
Some people might argue that valence zero does not exist because valence zero just means ‘ĝi is implicit’. That’s okay. I consider a different model here.
-
{agent/patient, agent/patient or more, valence, (in-)transitive} are models to explain how entities relate to the verbs
The curious case of igi/iĝi
Well, various verbs take an -igi or -iĝi suffix. More specifically, the following statements are true:
-
Some verbs only exist in their pure form (e.g. aparteni).
-
Some verbs only exist with an -igi suffix (e.g. listigi).
-
Some verbs only exist with an -iĝi suffix (e.g. ellitiĝi).
-
Some verbs exist in the forms {pure, -igi} (e.g. ŝajni, ŝajnigi).
-
Some verbs exist in the forms {pure, -iĝi} (e.g. levi, leviĝi).
-
Some verbs exist in the forms {-igi, -iĝi} (e.g. publikigi, publikiĝi).
-
Some verbs exist in the forms {pure, -igi, -iĝi} (e.g. koncentri, koncentrigi, koncentriĝi).
Let me sum it up: All combinations are possible. But I won’t claim that all dictionaries agree. Whereas the previous items might have slight differences between dictionaries, the last one is uncommon. The words koncentri, koncentrigi, and koncentriĝi exist in PReVO and reta-vortaro which use the same dataset (but apparently different versions because differences can be found in e.g. publiki), but PIV and glosbe deny the existence of -igi.
Sources
When mentioning dictionaries, I refer to these four:
As you can see, the first two dictionaries mention whether verbs are transitive (abbreviated as tr) or intransitive (abbr. ntr). So you have some clue about its valence. I chose ‘montri’ specifically, because it is an example verb which can be used transitive and intransitive.
Esperanto examples
Let us look at some examples:
-
‘levi’ is translated as ‘to lift’.
“Mi levas la libron” means I act upon the book by lifting it. So it has valence 2 and is therefore a transitive verb. -
‘leviĝi’ is translated as ‘to rise’.
“La birdo leviĝas” means the bird rises (less figuratively, it flies into the sky). It has valence 1 and is therefore an intransitive verb. -
There is no word ‘levigi’.
-
‘ĝeni’ is translated as ‘to bother’.
“Mi ĝenas mian edzinon” means I (as agent) am an inconvenience to my wife (as patient). It has valence 2 and thus ‘ĝeni’ is transitive. -
‘ĝeniĝi’ can be translated[1] as ‘to be bothered’.
“Mi ĝeniĝas” means that I am bothered (where it is not known by whom). -
‘pravi’ is translated as ‘to be right’
“Vi pravas” means that you are right. It has valence 1 and thus ‘pravi’ is nontransitive. -
‘pravigi’ is translated as ‘declare to be right’ or ‘show to be right’
“Nun vi pravigu vian kalkulrezulon” means that now you (as agent) should show that your calculation result (as patient) is correct. It has valence 2 and is transitive. -
‘praviĝi’ can be translated[1] as ‘to be corrected’ or ‘to be turned right’
“Per eraro la faro ne praviĝas” means the act (as patient) cannot be turned right with an error. The error reference is given by a preposition ‘per’ and hence we have valence 1 and it is intransitive. -
‘enliti’ does not exist and ‘enlitigi’ only exists in the glosbe dictionary, translated as ‘to put to bed’. Maybe one could formulate a sentence like “la patro enlitigas la filinon” (en. the dad puts his daughter to bed), but other dictionaries deny it.
‘enlitiĝi’ is translated as ‘to go to bed’. -
“Adam frue enlitiĝis pro la laboro” means Adam (as agent) went to bed early because of the work. It has valence 1 and thus is intransitive.
‘aparteni’ is translated as ‘to belong’. -
“La gratuloj apartenas al la partoprenantoj” means the wishes belong to the participants. It has valence 1 and is intransitive. Maybe? It turns out that ‘aparenti’ may only be used with preposition ‘al’. Usually we have the intuition that all prepositional phrases are optional, but few verbs like aparteni may only be used with this preposition. So maybe it has valence 2, but not in the sense of transitiveness because it has no object.
The following universal statements can be made:
-
Given a verb, the -igi suffix adds another entity to the sentence structure. The additional entity increases the valence by one.
-
Given a transitive verb, the -iĝi suffix removes some entity because the transitive sentence structure ‘agent verb patient’ corresponds to ‘patient verb-iĝi’. It decreases the valence by one.
-
Weather-related verbs never have -igi or -iĝi forms.
-
Any description with valence or transitiveness breaks once we encounter a verb like aparteni which requires a subject and a pri-preposition. Valence 2 usually means subject-and-object, but here it means something different.
The following incomplete statements or ideas can be expressed:
-
For igi-verbs, the additional entity has the notion of “initiating/instigating an action”. In English, the -igi verb would be expressed as “to be made to” or “to be instigated to”. But this English explanation does not fit e.g. for ‘pravigi’ where the subject does not instigate the object but the subject has to show the object.
-
If some transitive verb describes that the agent performs an act using the patient, the verb with -iĝi describes that the patient is in the state of the described act. In English, this is similar to passive voice.
-
If some verb and its counterpart with prefix mal- exist[2], they are usually used with the same structure. This is rarely true for other affixes.
How verbs shall be taught in my opinion
I do believe I am interested in the topic, because I just learned verbs in the wrong way autodidactically. Whereas English teachers in high school showed me patterns how to use verbs, I mostly studied translations for Japanese and Esperanto. In Anki, I can find plenty of examples:
In summary, there are verbs used like …
-
‘verb’ which is limited to weather-related statements in Esperanto.
-
‘A verb’ where A is an agent followed by optional ‘preposition C’ patterns in any ordering.
-
‘A verb B’ where A is a patient followed by optional ‘preposition C’ patterns in any ordering.
-
‘A verb B’ followed by optional ‘preposition C’ patterns in any ordering.
-
‘A verb preposition B’ followed by optional ‘preposition C’ patterns where preposition phrases have any ordering. Unlike C, B is required.
Now it would be completely wrong to claim that the semantics of the verb together with the sentence structure exists equivalently in every other language. Hence it does not suffice to study its translation alone. It neither suffices to know which valence a verb has or more trivially whether it is transitive or intransitive. Instead I should study a verb together with examples for the sentence structures which exist in Esperanto. This way one should be able to engrave into one’s brain the structures a verb may be used with.
Conclusion
-
Dictionaries do not agree about usages of verbs. I do believe because first, it is difficult to cover how people use language (e.g. is “Ich ende” a valid sentence in German?) and second, the very strong multilingual influence for Esperanto leads to new (often superfluous) patterns all the time.
-
Subject and objects are implicit arguments for a verb. Other arguments are provided with explicit prepositions which already give a hint how they relate to the verb.
-
The concepts of valence and (in)transitiveness and do not cover the true complexity of verb usage.
-
Verbs in -igi forms usually mean that the subject instigates the action. Verbs in -iĝi form usually mean that the subject is passively in a state and it does not state who is acting upon the subject.
Finally, I understood this topic good enough!