Clauses

21.1 Simple Sentences and Independent Clauses

The unmarked word order in a simple sentence in Alashian is SVO. Adverbs, including expressions of time and place as well as indirect objects, typically go between the verb and direct object, though clause-final position is possible for emphasis or when the adverbial phrase is too heavy (i.e., any more complicated than a preposition + noun).

VSO and OSV orders are also possible in order to emphasize the verb or direct object respectively, but some with some syntactic quirks. Fronted verbs are never accompanied by preposed clitic pronouns, meaning subject clitics are not used and object clitics must follow the verb. When the direct object is fronted, the verb must be accompanied by a resumptive object clitic (as though the object had been removed from the sentence). Adverbs tend to be clause-final whenever VSO or OSV order is used.

VSO order is, however, the unmarked order in older texts such as the Bible. The same rules apply as above.

Note that clitic pronouns are treated as part of the verb phrase, not as autonomous noun phrases. The aforementioned word ordering rules only apply for nouns and full-form pronouns, never clitics. The genitive components of perfect verbs, however, pattern as true subjects.

21.2 Compound Sentences and Conjunctions

Compound sentences consist of two or more independent clauses joined by a coordinating conjunction. No word order modifications are required when two simple sentences are joined into a compound sentence other than the insertion of a conjunction in between the two.

The conjunction ve- (spelled υε-) means “and”, and is used to conjoin both noun and verb phrases. It is a clitic conjunction, and so is written as part of whatever word immediately follows it. It becomes va- (υα-) when immediately followed by /h/ or /ʁ/, with the exception of the definite article ha-, where it remains ve- (likely a dissimilatory phenomenon). In formal usage the form va- is also used immediately before the reduced form of the definite article n- (υανήν vanēn “and the eye”), though in practice ve- is heard here in most spoken usage (υενήν venēn).

The non-clitic conjunction τζε če is also frequently heard for “and”, and exists in free distribution with ve-. It is a direct borrowing of Greek και ke “and”. ¹

The conjunction ιυ yu means “or”, both in noun phrases and in verb phrases. Its actual pronunciation varies significantly, with all of /ju ~ u ~ uː ~ ɑː ~ wu ~ wuo/ being heard among different speakers.

The conjunction με me (from Greek μα ma) means “but” or “whereas”, and is used to express contrast.

The reduplicated conjunction ιν... ιν... 'in... 'in... means “either... or...”. Its negative counterpart is βλω... βλω... blā... blā... “neither... nor...”.

If both independent clauses share a single subject, it need only be explicitly present in the first clause. Clitic pronouns, however, can never be dropped in this fashion. Similarly, if both clauses share a direct object, it may be mentioned only in the first clause, with an accusative clitic pronoun taking its place in the second clause.

21.3 Relative Clauses

The relative clause is one means of joining a subordinate clause to an independent clause, by converting the former into a modifier of a noun phrase in the latter. In such a context, the subordinate clause will often be called the embedded clause, while the independent clause is termed the matrix clause. Alashian subordinate clauses display a number of distinct syntactic behaviors and word orderings not seen in independent clauses.

Alashian contrasts restrictive and non-restrictive relative clauses, a contrast much like the formal English usage of 'that' and 'which'. Restrictive relative clauses limit the scope of their head (as in “My brother who lives in Athens is named Alexandros”, which has a narrower sense than just “my brother”), while non-restrictive relative clauses behave more like appositives and simply provide supplementary information (as in “My brother, who lives in Athens, is named Alexandros”).

21.3.1 Restrictive Relative Clauses

Restrictive relative clauses are formed using the complementizer δε de “that, who”, which always follows the noun phrase in the matrix clause that it modifies, and precedes the embedded clause. It is not a relative pronoun as seen in many Indo-European languages, but simply a conjunction that links a complete subordinate clause to the outer matrix clause. Consequently, resumptive pronouns are frequently seen in the embedded clause, so that noun phrases such as “the school that I attended” and “the restaurant in which we ate” are rendered “the school that I attended it” and “the restaurant that we ate in it”. In practice, however, resumptive pronouns are rarely used if they are the subject of the embedded verb, and optional if they are the direct object; they are mandatory in all oblique positions.

As can be seen above, the word order of a subordinate clause differs from that of an independent clause. The primary order is VSO, with nominative clitics always omitted from the verb phrase and accusative clitics always following the verb. However, embedded clauses involving the verb “to be” retain their usual word order.

The subordinating conjunction δε de will often reduce to just δ' d when the following word begins with a vowel or the definite article. This is nearly universal in speech, and is optional in all written registers.

21.3.2 Non-Restrictive Relative Clauses

Non-restrictive relative clauses are introduced by proximal demonstrative pronouns, namely 'aḏḏekā (masculine singular), 'aḏḏičī (feminine singular), and 'aḏēlek (plural), agreeing in gender and number with their head noun in the matrix clause.

Unlike with restrictive relative clauses, no resumptive pronouns may be used in non-restrictive clauses, a consequence of the presence of an actual pronoun as opposed to a generic subordinator such as de. This also means that the accessibility hierarchy becomes a concern in non-restrictive clauses—only the subject and direct object of the embedded clause are accessible from the matrix clause.

Other nouns in the embedded non-restrictive clause may be accessed by inserting the subordinating conjunction de after the demonstrative pronoun and using resumptive pronouns, giving 'aḏḏekā de, 'aḏḏičī de, and 'aḏēlek de. Some more conservative dialects may use these forms in all non-restrictive clauses, including those previously described.

21.4 Substantive Clauses

21.4.1 Object Clauses

Object substantive clauses are dependent clauses that serve as the direct object complement of a particular transitive verb. In Alashian such clauses are introduced by the complementizer με me.

Με me also serves to introduce the complement of an impersonal adjective.

Secondary predicates (known also as 'extended objects' in Alashian), marking the state of a direct object when the main predicate verb takes place, are akin to object complement clauses. In older literature and more conservative dialects, they are formed with the structure and-complementizer verb. In modern speech a non-finite construction is more common: and-by infinitive.

21.4.2 Predicate Clauses

Predicate clauses are a subclass of substantival clauses that consist of a dependent clause embedded within the complement of a copula. They are introduced with μώ δε/δ' mā de/d, which is more or less functionly equivalent to English “that which”.

21.5 Adverbial Clauses

21.5.1 General Structure of Adverbial Clauses

Adverbial clauses are dependent clauses that function as an adverb within the matrix clause, specifying various additional attributes of the main verb. As with other dependent clauses, the adverbial clause is introduced by a subordinating conjunction. However, if the adverbial clause is fronted and appears at the start of the matrix clause, a special “resumptive” conjunction is inserted to indicate the end of the adverbial clause.

The use of conjunctions in many ways parallels the use of “if” and “then” in English, with “if” being a mandatory conjunction introducing a dependent clause and “then” being an optional conjunction introducing the main clause. As in English, the resumptive “then” is only used if the dependent clause precedes the main clause (i.e., “if X then Y”, never “then Y if X”). However, unlike English, the resumptive conjunction is never optional and is required in many more circumstances than in English.

21.5.2 Clauses of Time

Clauses of time indicate the chronology of the action in the main clause relative to the action of the adverbial clause. All clauses of time make use of the resumptive conjunction τυώτε tuote, which frequently shortens to τυώτ' tuot when followed by a vowel sound.

The distinction between λενώ lenā and μετώ ιυνδε metā yunde, both translated as “after”, is generally one of causality. Lenā implies that the subsequent action is either a direct consequence of the preceding action or immediately follows it within a single reference frame; metā yunde suggests a greater disconnect, with no necessary causal connection, and the possibility of intervening events between the two actions.

21.5.3 Clauses of Place

Clauses of place indicate the location where a particular event happened. The resumptive conjunction is usually θών ṯān “there”, although conjunctions indicating direction rather than static location may optionally use forms such as αδών 'adān “to there” or θώνα ṯāna (a now otherwise-defunct adverb originally meaning “to there”).

As can be seen in the table above, the majority of place conjunctions are based on the noun μακκούν məkhūn “place” plus a relative conjunction, so that “wherever”, for example, is literally expressed as “in each place that”. However, in the spoken language, it is becoming increasingly common to use the short forms δ' ηαλού d halū “wherever” and δ' ιλού d 'ilū “to wherever”, consisting of declined forms of the prepositions ηαλ hal “on” and ιλ 'il “towards”. Informally these may even be written/pronounced as δώλυ dālu and δείλυ dīlu respectively.

21.5.4 Clauses of Manner

Clauses of manner indicate how the action in the main clause was performed, either by elaboration or by comparison. The resumptive conjunction used with such clauses is κάκ kak “thus”, except for χήν xēn “as if”, which uses φα/φ' fa/f.

21.5.5 Clauses of Cause

Clauses of cause indicate the reason or motivation behind an action. The resumptive conjunction for such clauses is φα fa, or φ' f when followed by a vowel.

For the most part, all of the above forms are interchangeable as far as semantic considerations are concerned. The main difference is one of register, with kad being the most neutral, mā being rather colloquial, and the phrasal forms more typical of higher styles.

21.5.6 Clauses of Contradistinction

Conjunctions of contradistinction indicate that the action of the main clause is somehow unexpected or incomplete when the information in the adverbial clause is taken into consideration. The resumptive conjunction may be either φα/φ' fa/f or κακ kak in more or less free variation.

21.5.7 Clauses of Comparison

Clauses of comparison express a relation (in terms of 'more' or 'less') between two clauses—often, but not necessarily, with an adverb serving as the standard of comparison. The conjunctions used are kyu [adv] pra “more than” and lətfat [adv] pra “less than”. The resumptive conjunction, though rarely used, is φα/φ' fa/f.

Much of the comparison clause can be elided if the verb in both clauses is identical. If the comparison clause is reduced to just a single pronoun, disjunctive forms will be used.

The absence of an adverb results in a comparison of quantity.

When the standard of comparison is “better” or “worse” (the two adjectives with synthetic comparatives in Alashian), it is rendered as καλείττερατ kalītherat and σ̄ιρούττερατ širūtherat respectively. In other words, the adverbs are formed from the comparative stems. However, unlike adjective comparison, it is not unusual for the adverb κιυ “more” to remain present.

21.6 Conditional Sentences

The conditional sentence consists of two clauses, the protasis (the condition or 'if' clause) and the apodosis (the result or 'then' clause). As with adverbial clauses, the protasis must always be introduced by a conjunction and the apodosis only if it follows the protasis, with the protasis acting syntactically as a dependent clause and the apodosis as an independent clause.

21.6.1 Factual Conditionals

A factual conditional sentence is one in which the truth of the protasis is unknown, and it simply follows that if the condition in the protasis is true, then the resulting state described in the apodosis has or will come to pass. The protasis is introduced by the conjunction μίρ mir “if” and the apodosis by φα fa “then”, which becomes φ' f when followed by a vowel. Since the clause introduced by φα is independent, clitic pronouns will generally appear in their usual pre-verbal position.

The verb in the protasis may appear in the present or present perfect; the present tense form is also used with future meaning, while the present perfect is used for all past tense meanings. The verb in the apodosis is unrestricted and may appear in any indicative mood tense. Notice that a perfect or pluperfect verb in the protasis generally does not need the adverb ple seen in most other positions.

The apodosis may also appear in the volitive as described previously in section 19.3.3 when the action both takes place in the future and is viewed by the speaker in a positive light.

The factual conditional structure, with both the protasis and apodosis in the present tense, is also used for universal conditionals, which state that the result clause always has and always will follow from the condition. In such cases μίρ mir can often be translated as “whenever” or “every time” in addition to “if”.

When the verb of the protasis is negated, the conjunction mir and the negative lā fuse into a single word ιλλώ 'illā “if not”. ²

21.6.2 Counterfactual Conditionals

A counterfactual conditional sentence describes a hypothetical state. The condition in the protasis is untrue or unlikely, and the result clause describes what supposedly would have happened if the condition had been true. The protasis is introduced by the conjunction λού lū “if” and the apodosis by φα/φ' fa/f “then”.

The protasis may appear in the present, present perfect, or pluperfect, while the apodosis may appear in any indicative mood tense.

When the verb of the protasis is negated, the conjunction λου lū and the negative λω lā fuse into a single word λυλλώ lullā “if not”.

21.7 Interrogative Sentences

21.7.1 Direct Questions

Interrogative sentences can be formed in three ways: with an interrogative pronoun, with an interrogative particle, or with interrogative word order.

Interrogative pronouns (as described previously) generally appear at the start of the sentence, no matter their usual syntactic role. If multiple interrogative words are present in a single sentence, generally only one is fronted, while the other remains in situ.

Yes/No questions can be formed either with the interrogative particle α 'a (dialectically ηα ha) or a change in word order. When the interrogative particle is used, it appears in absolute clause-initial position, while the rest of the clause appears in its normal indicative-mood order. This is accompanied by a rising intonation.

Alternatively, if 'a is not used, interrogative mood can also be marked by a change to a VSO (verb-subject-object) word order alongside rising intonation.

21.7.2 Indirect Questions

Indirect questions only require a conjunction if the entire clause is being questioned. In this case, the conjunction αν 'an “whether” is used, and the question appears in a dependent clause.

If the question contains an interrogative pronoun of some sort, it can be incorporated as-is with no conjunction at all.