Introduction

This book describes the major phonological, orthographic, syntactic, and semantic properties of Najan, my first and so far only conlang. Najan is an a priori, analytic, isolating conlang featuring perfectly phonemic orthography, an unambiguous context-free grammar, and pronoun binding and quantification rules inspired by first-order logic. The Najan name for Najan is kʊ naj vɪx [kʰʊ näʒ vɪɾ̥ʰ].

I have no formal linguistic training, so the language is not very naturalistic or elaborate as far as conlangs go. Nevertheless, it’s been a joy to work on, and I’m happy with how it has evolved and am excited to share it with others.

Soli Deo gloria.

Phonology

Consonant clusters with a single point of articulation are geminated. For instance, a word-final k followed by a word-initial g would be pronounced [kːg] (note the lack of the usual aspiration) since both are velar. In contrast, the ⟨k⟩ in ⟨-k t-⟩ is aspirated since ⟨t⟩ is alveolar.

A plosive-fricative cluster with a single point of articulation becomes an affricate. t and d, which normally denote alveolar plosives, shift their point of articulation before dental or post-alveolar fricatives to assimilate with the fricative.

The clusters tl and dl become the lateral affricates [t͡ɬ] and [d͡ɮ], respectively. Similarly, in some dialects, sl and zl become the lateral fricatives [ɬ] and [ɮ].

Phonotactics

Najan is monosyllabic, with C(C)V(C)(C) structure. Within an onset or coda, any plosives, fricatives, and taps must all be voiced or all be voiceless. For example, ⟨ts-⟩ and ⟨tl-⟩ are valid, but ⟨tz-⟩ is not.

Stress

Najan uses a stress accent system, with stressed words receiving greater volume, a longer vowel, and higher tone. Najan is also stress-timed, with stressed words occurring at relatively constant intervals. Typically, verbs are stressed and particles unstressed, except for emphasis.

Orthography

In the table below, each Najan glyph is labeled with the transliteration of its grapheme’s Najan name (in angle brackets), along with its most common associated phone (in square brackets). Each grapheme name begins with the grapheme itself. Since Najan is perfectly phonemic, this table also represents all the Najan phonemes. Note however that these phonemes are not always pronounced exactly as the example phones, due to the allophonic variations described in Phonology.

Sentences are written left-to-right, then top-to-bottom. Najan has no punctuation except spaces between words.

Syntax

Najan has the following word classes:

• Verbs (fyas)
• Determiners (qa saθ dan, means of instantiation)
• Pronouns (nʊ zax kʊv, borrowed identity)
• Prepositions (qa qac dan, means of relation)
• Conjunctions (qa cob vom, logical operator)
• Particles (lʊg)

Najan has no nouns, adverbs, or adjectives. Verb phrases, pronouns, and determiner phrases can all function as noun phrases. (A verb phrase may be used as a verbal noun phrase without any inflection or other marking.) Prepositional phrases fill the roles of adverbs and adjectives.

Formal Grammar

Najan is a deterministic context-free language whose formal grammar can be expressed in extended Backus-Naur form as follows:

clause ⇒ ⟨mood ᴘᴛᴄʟ⟩ ⟨VP⟩
| ⟨PP⟩ ⟨clause⟩
| ⟨restrictive clause⟩ ⟨clause⟩
| ⟨non-restrictive clause⟩ ⟨clause⟩

ᴘʀᴇᴘ phrase (PP) ⇒ ⟨ᴘʀᴇᴘ⟩ ⟨NP⟩

restrictive clause ⇒ nʊ ⟨VP⟩

non-restrictive clause ⇒ vew ⟨VP⟩

verb phrase (VP) ⇒ [ ⟨aspect ᴘᴛᴄʟ⟩ ] ⟨verb⟩
| ma ⟨VP⟩
| ⟨PP⟩ ⟨VP⟩
| ⟨restrictive clause⟩ ⟨VP⟩
| ⟨non-restrictive clause⟩ ⟨VP⟩
| txa ⟨VP⟩
| ⟨ᴄᴏɴᴊ⟩ ⟨VP⟩ ⟨VP⟩

noun phrase (NP) ⇒ ⟨VP⟩ | ⟨restricted NP⟩

restricted NP ⇒ ⟨ᴘʀᴏ⟩
| ⟨ᴅᴇᴛ phrase⟩
| ⟨quotation⟩
| ⟨non-restrictive clause⟩ ⟨restricted NP⟩
| txa ⟨restricted NP⟩
| ⟨ᴄᴏɴᴊ⟩ ⟨restricted NP⟩ ⟨restricted NP⟩

ᴅᴇᴛ phrase ⇒ ⟨ᴅᴇᴛ⟩ [ ⟨ᴘʀᴏ⟩ ] ⟨VP⟩

quotation ⇒ tca ⟨quoted word⟩+ tca

quoted word ⇒ ⟨word⟩ - tca - tsa | tsa tca | tsa tsa

Modality

Every clause in Najan begins with a mood particle expressing how the speaker relates the clause to the world or to other possible worlds. The mood particles are clipped from corresponding verbs, e.g. je (sensory indicative mood) from jec (sense, perceive, or intuit).

Indicative Moods

The indicative moods are used to assert a claim about how the world actually is. These moods are distinguished by evidentiality, i.e. the speaker’s source of information supporting the claim.

The sensory mood (sɴs) is for claims supported by the speaker’s own direct sensory experience or intuition.

The deductive mood (ᴅᴇᴅ) is for conclusions the speaker has deduced rationally from other accepted evidence. The conclusion could be based on sense data, hearsay, or other inferences, as long as the conclusion itself comes from the speaker’s own reasoning. Use of the deductive mood doesn’t imply that the speaker has consciously constructed a formal logical proof for their claim, but it should imply that they believe they could sketch a convincing informal argument on request. Hunches should be expressed using the sensory mood, not the deductive.

The reportative mood (ʀᴘʀᴛ) is for claims originating from someone other than the speaker, such as something the speaker has heard or read.

The choice of indicative particle by itself has no bearing on how strongly the speaker believes the claim. For instance, it’s entirely possible to be more convinced of a reportative statement originating from a highly reliable and authoritative source than of a sensory statement based on a vague feeling or unclear view. The speaker may convey how strongly they believe their own statement using an epistemic mood particle.

Interrogative Moods

The interrogative moods are for asking different types of questions.

A content question (ᴄᴏ̨) requests missing information using the interrogative pronoun vel. The response should supply the missing information that would make the complete clause true. To this question…

…I could respond je djan (It’s Jon).

Informally, one can omit vel, if it’s apparent what’s being asked: gwi kʊv (Name?).

A content question may request multiple pieces of information.

This questioner has asked for two pieces of information here: who did it and what they did. Note that vel is a pronoun, so it can’t stand in for a verb directly. However, using pɪn (do) as the verb allows asking about actions, by using vel as a causal-final argument.

A polar question (ᴘᴏ̨) prompts for a yes-or-no answer, confirming or denying the clause.

The reponse can usually be very brief, e.g. je ksi (it’s true) or je gle (it’s false).

An alternative question (ᴀᴏ̨) should pose a series of mutually exclusive options, only one of which makes the overall clause true. Typically this involves an exclusive logical disjunction. It’s possible but very uncommon for an alternative question to contain multiple sets of options, prompting multiple selections from the respondent. A question with two opposite possibilities would typically be expressed as a polar question but could be expressed as an alternative question for emphassis: are you or are you not my people?

Any question may be predicated on faulty premises, whether implicit or explicit. In that case, the respondent may choose to contradict these assumptions rather than answer the question on its own terms.

Deontic Moods

The deontic moods are for asserting how the world shall be or ought to be.

The commissive mood (ᴄᴍs) indicates that the speaker intends to bring about the clause so that in the future it will be become or remain true.

The imperative mood (ɪᴍᴘᴇʀ) expresses a request or command for the listener to make or keep the clause true.

The volitive mood (ᴠᴏʟ) expresses a wish or desire of the speaker. Unlike the commissive or imperative moods, it does not imply that the speaker or listener is responsible for actualizing that wish.

Hypothetical Mood

The hypothetical mood is for expressing counterfactuals. A hypothetical clause posits that though the claim is not actually true, it would be true under other conditions. Clauses in this mood generally involve a conditional argument to clarify under which counterfactual circumstances the claim would be true. Note that the conditional argument should apply to the verb phrase within the hypothetical clause, not to the entire hypothetical clause itself.

Conditional claims where the consequent could actually be or become true should instead use the indicative mood with a conditional argument to the verb.

Modal Clauses as Truth Propositions

It’s possible to analyze clauses in any mood as corresponding to assertions about what is actually true. Though this correspondence is most obvious for the indicative moods, it’s true for the others as well. For instance, a verb phrase X belonging to a clause in the volitive mood asserts the proposition that the speaker wishes for X to be true. Likewise, a clause in the polar interrogative mood signifies that the speaker wonders whether X is true and would like the listener to provide an answer.

This kind of analysis is unusual but can be useful for correctly interpreting logical operators or prepositional phrases applied to clauses. For example, je txa gleð means I sense that it’s not raining, whereas txa je gleð means I don’t sense that it’s raining, which is a weaker claim. Similarly, nʊ zʊm kca vɪg literally means regretfully, I request that you go, whereas kca nʊ zʊm vɪg means I request that you regretfully go.

Relative Clauses

A relative clause modifies a verb phrase, noun phrase, or clause using a verb phrase complement whose implicit subject is the modified head.

A restrictive clause, beginning with nʊ (that), restricts its head to only those instances where the verb phrase complement is true (again, using the head as the subject).

Here, the relative clause nʊ tir helps define the entity the speaker is referring to: one belonging to the set of excellent dogs. Perhaps there are also some not so good dogs under consideration. The use of the restrictive clause rules these dogs out as referents of this noun phrase.

A non-restrictive clause, beginning with vew (which), provides additional information about the head but does not narrow down the set of possible referents. In other words, it describes the head without defining it.

Here, vew tir asserts that all dogs are good. We could remove the non-restrictive modifier entirely, and the meaning of gigi ⟨gi⟩ — thedefinite determiner ŋix would not change.

Restricted noun phrases, such as determiner phrases and pronouns, cannot be further modified by restrictive clauses or prepositional phrases. For instance, nʊ tayn ko (happy I) is ungrammatical. Conversely, non-restrictive clauses can modify restricted noun phrases.

In this example, vew tir does not contribute to the identification of the dog, nor does it describe dogs in general. It provides extra information about a particular dog that has already been identified.

Prepositional Phrases

A prepositional phrase comprises a preposition and its noun phrase complement, assigning that complement some semantic role within the clause or verb phrase that the prepositional phrase modifies.

Prepositions are derived from corresponding verbs by clipping. For example, the preposition cici ⟨ci⟩ — bymarks the complement as an agent marks its object as an agent and is derived from the verb cim, agent.

Subjects & Relations

Most prepositional phrases—and the verb arguments they express—have relatively fixed definitions, so if one knows the meaning of a verb, then the meaning of its arguments follows straightforwardly. Two semantic roles are exceptions to this pattern of predictability: the subject (ŋɪθ) and relation (qac). These roles are essential components of the semantics of each verb.

Every Najan verb describes some state of being. The subject of a verb, marked using the preposition ŋɪ, denotes the entity that is in that state. (It can be very difficult to provide a good gloss for this preposition since in English the subject is positional. Therefore, I will generally gloss ŋɪ as just “sʙᴊ”.)

Many verbs also permit a secondary relational argument, marked using qa, which most often means something like of, with respect to, with regard to, or in relation to.

For example, in the verb zim (want), the subject is the one who is experiencing desire, and the relation is that which is wanted.

For a given Najan verb, the meanings of ŋɪ and qa may or may not align well with the subject and object of a similar English verb. As a general rule, the entity most strongly pertaining to or affected by the verb will be defined as its subject. Taking for example the verb tcet (be eaten), the subject is the thing consumed, and the relational argument is the consumer. These roles are exactly reversed for the related word hɪm, eat.

Causes & Conditions

An agent (cim) of a verb, marked by the preposition cici ⟨ci⟩ — bymarks the complement as an agent, is its animate, voluntary cause.

In contrast to an agent, a cause (slon) is involuntary. The preposition slo translates to because of.

A cause can also be inanimate.

When the complement of a causal argument is a verb phrase, slo functions like the English subordinating conjunction because.

The causal-final (ғɪɴ) preposition tce identifies a final cause: a purpose, outcome, or product of some state. It can translate to for or so that. A final cause may be a verb phrase or noun phrase.

The conditional preposition xo specifies a condition or dependency required in order for the modified expression to apply. It most closely translates to depending on.

When the complement is a verb phrase, the conditional functions like the English subordinating conjunction if.

Time & Place

Najan has four prepositions for characterizing time and place:

1. lɪ for location at or within a point or region of space or time
2. we for movement to or into a point or region
3. xu for movement from, away from, or out of
4. zi for movement through, throughout, or across

Each of these prepositions is inclusive; i.e. we includes the destination, xu includes the origin, and zi includes the entirety of the path, region, or duration.

The complements of these prepositions may be explicit places or times, but they could also be something else, such as a person or event. In the latter case, the argument refers to the spatio-temporal location of that object; e.g. lɪ ko (literally at me) means at my time and place.

English supplies a plethora of prepositions for describing spatio-temporal relations. For instance, a chair may be at, on, under, or beside a table. In Najan, each of these relations would use the same (locative) preposition: zaza ⟨za⟩ — a; anindefinite determiner lɪ zaza ⟨za⟩ — a; anindefinite determiner cats qa muð dan, a chair at a table.

The preposition alone is often sufficient in context to convey the full relationship, but to add clarity or expressiveness, one can replace the complement of the preposition with a more specific expression. For example, since the preposition we (to) is always inclusive of the specified destination, we gigi ⟨gi⟩ — thedefinite determiner vatθ (to the end) means that the expression holds true up to and including the end. To convey that the expression held true up to (but not necessarily including) some time before the end, one could say we zaza ⟨za⟩ — a; anindefinite determiner qa gigi ⟨gi⟩ — thedefinite determiner vatθ kle, lit. to a past [time] of the end.

Similarity & Comparison

The semblative (sᴇᴍʙʟ) preposition marks something the complement is similar to. The semblative particle kɪ means as, like, or in the manner of.

The comparative (ᴄᴏᴍᴘ) preposition marks a point of comparison or reference within a verb phrase of degree.

The comparative particle ve means roughly than or in comparison to, but unlike these English terms, the comparison is always against the subject of the verb. This avoids ambiguous comparisons such as I like cheesecake more than you—more than you like cheesecake or more than I like you?

A comparative argument can be combined with the distributive determiner su to form a superlative:

Verb Phrases

Arguments

Arguments to a verb are expressed as optional prepositional phrases. As a result, arguments have no fixed order, and a single verb can take multiple arguments of the same semantic role.

In the following, you and I are both eating the same fruit:

Reversing the order of the arguments usually does not affect the meaning:

However, sometimes order does matter, for instance when universal and existential quantification are involved.

In the next example, the ball and glove are thrown as part of the same action:

Since arguments are simply modifiers, they are not syntactically required and may be omitted if they can be inferred from context. When asked who I am, I could reply:

Or, since I is the expected subject:

Tense & Aspect

Tense specifies when along a timeline a verb phrase occurs (e.g. past, present, future), whereas aspect characterizes how the verb phrase extends over time. In Najan, aspect is marked syntactically with a (possibly null) particle immediately preceding the verb or verb phrase. In contrast, tense has no direct syntactic representation, so it must be marked explicitly with temporal arguments, inferred from the surrounding context, or else assumed from the verb’s aspect.

The continuous (ᴄᴏɴᴛ) aspect views a verb as ongoing or in progress. It is the default aspect, in the absence of an aspect particle. Without temporal arguments or additional context, a continuous verb phrase may be assumed to be in present tense. Najan does not distinguish between continuous and progressive aspects.

The perfective (ᴘғᴠ) aspect views a verb in its entirety, as a complete action or state. Without other clues, a perfective verb phrase is most likely to be in past tense. For example, je cici ⟨ci⟩ — bymarks the complement as an agent ko θe tcet is more likely to mean I ate (simple past) or I have eaten (past perfect) than I will eat (simple future).

The inchoative (ɪɴᴄʜ) aspect views a verb as beginning, as in je dje sic, [he] starts to cry. Inversely, the cessative (ᴄᴇss) aspect views a verb as ending: je vat sic, [he] stops crying. By default, verbs in these aspects are in present tense.

The habitual (ʜᴀʙ) aspect views a verb as occurring regularly over some period of time, centered on the present by default. If je cici ⟨ci⟩ — bymarks the complement as an agent to vɪg is in present tense, it means you are going, and the corresponding habitual phrase, je cici ⟨ci⟩ — bymarks the complement as an agent to fxe vɪg, means you go [regularly]. If we instead assume past tense, then they translate respectively to you were going and either you would go or you used to go.

The gnomic (ɢɴᴏ) aspect views a verb as a general truth with little or no temporal structure. As such, the concept of tense is usually inapplicable. Gnomic verb phrases often feature generic determiner phrases as arguments and express an aphorism, a mathematical truth, etc.

The gnomic verb phrase above expresses a general mathematical fact. If it were written in the continuous aspect instead (without hon), it would leave open the possibility that two might not have always been or might not always be greater than one.

Inferring Tense from Context

As described above, aspect markers can also carry a small amount of tense information. For example, perfective aspect often implies past tense. However, these associations are secondary to context. Suppose a speaker first says:

By using the preposition lɪ, the speaker has established a context in the future. Suppose the speaker then says je cici ⟨ci⟩ — bymarks the complement as an agent ko θe tɪz. In isolation, since this sentence is in perfective aspect, one might interpret it as I have spoken (present perfect) or I spoke (simple past). However, since future tense has already been established, one should instead translate the second clause as I will speak or I will have spoken (sometime tomorrow, after having gone).

Noun Phrases

Any verb phrase can also function as a noun phrase, signifying an abstract state or action (e.g. being human, running, love). To talk about concrete instances, sets, or amounts of something (e.g. you, the victory, two dogs, some water, several words), we can use pronouns or determiner phrases.

Pluralization

A verb phrase can be pluralized by preceding it with the particle ma. Precisely, ma followed by a state means to be a set of entities that are all in that state. This set may contain any number of elements, including zero or one.

Unlike plurals in English, instantiating the ma plural form requires the use of a determiner. The following attempted translation of I hear people over there is incorrect (and nonsensical):

Adding the indefinite determiner gives the intended meaning:

Pronouns

Najan has two first-person plural pronouns, distinguished by clusivity. The inclusive pronoun djof includes the speaker, those addressed, and possibly one or more third-person referents. The exclusive pronoun jal includes the speaker and at least one third person but excludes those addressed.

The interrogative pronoun vel stands for something unknown and is essential to the formation of content questions, where it indicates the missing information that the asker wishes to know.

The assignable pronouns jow, með, paŋ, and tsut have no fixed properties, such as person, number, or gender. They can be bound (or rebound) to any unquantified noun phrase, via determiner phrases.

Determiner Phrases

A determiner phrase consists of a determiner, an optional pronoun to bind, and a verb phrase complement. The phrase signifies some entity such that the complement holds true with that entity as the subject. Najan has just four determiners:

The indefinite determiner zaza ⟨za⟩ — a; anindefinite determiner introduces a new specific instance of the complement. It corresponds to existential quantification in logic.

The definite determiner gigi ⟨gi⟩ — thedefinite determiner refers to a known instance of the complement. The referent should be unique (e.g. a proper noun) or otherwise obvious from context. There is only one entity that is the sun:

We can express a demonstrative (this or that) by modifying the complement of a definite determiner with nʊ tlɪ (here) or nʊ dwa (there).

The distributive determiner su expresses that the clauses containing the determiner phrase (and its pronoun, if any) hold true for each instance of the complement. It corresponds to universal quantification in logic.

The generic determiner ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense refers to the complement in a general, nonspecific sense. It’s often used with the gnomic aspect. This determiner doesn’t map cleanly to any one determiner in English; for instance, a man’s gotta eat, cats are cute, and the giraffe is an African species all have a generic subject.

In English, the indefinite articles a and an do not encode specificity. For instance, I’m looking for a book may or may not mean I have a specific book in mind. In Najan, the indefinite determiner is always specific:

The generic determiner is always nonspecific.

Pronoun Binding

Binding a pronoun in a determiner phrase allows referring back to it in a concise and unambiguous way.

Plural and Uncountable Determiner Phrases

The complement in a determiner phrase may be uncountable or plural, in which case the entity identified by the determiner phrase is a set or quantity.

In fact, Najan does not distinguish between countable and uncountable determiner phrases since determiner phrases always identify discrete entities. In English, rock is countable in a small rock but uncountable in a hunk of rock. In Najan, the word for rock, kut, means to be made of stone. Thus zaza ⟨za⟩ — a; anindefinite determiner kut, literally a thing that is made of stone, could be translated into English either as a rock (countable) or as some rock (uncountable).

Ordering Determiner Phrases

We have to be careful about the order of arguments when distributive determiners and indefinite determiners are both present since they are not commutative with each other.

It’s true that every human has a human mother:

It is however not true that all humans share a single universal mother:

Quotation

A quotation begins and ends with the particle tca and functions as a noun signifying the contained words themselves. The particle tsa is used within a quotation just before tca or another tsa to indicate that the second particle should be interpreted as part of the quotation, not as a particle within the sentence containing the quotation.

Note: Determiner phrases within quotations cannot affect the bindings of surrounding pronouns. Assignable pronouns contained in a quotation may have completely different referents than in the surrounding context—or even none at all, when using a quotation to talk about the pronouns themselves.

... kʊ naj vɪx

🔗 gigi ⟨gi⟩ — thedefinite determiner qa saθ dan

fo qa ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense jow nʊ lɪ kle qa jow θe tɪz ŋɪθ dje kʊv

🔗 zaza ⟨za⟩ — a; anindefinite determiner qa saθ dan

fo qa ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense jow nʊ txa lɪ kle qa jow θe tɪz ŋɪθ dje kʊv

🔗 ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense qa saθ dan

fo qa ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense nʊ ðus ŋɪθ dje kʊv

🔗 caycay ⟨cay⟩ — change; evolve loð

...

🔗 cici ⟨ci⟩ — bymarks the complement as an agent qa qac dan

...

English-Najan Dictionary

🔗 a zaza ⟨za⟩ — a; anindefinite determiner ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense

🔗 an zaza ⟨za⟩ — a; anindefinite determiner ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense

🔗 by cici ⟨ci⟩ — bymarks the complement as an agent

🔗 change caycay ⟨cay⟩ — change; evolve

🔗 evolve caycay ⟨cay⟩ — change; evolve

🔗 the gigi ⟨gi⟩ — thedefinite determiner ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense

Najan-English Dictionary

🔗 gigi ⟨gi⟩ — thedefinite determiner the

definite determiner

🔗 zaza ⟨za⟩ — a; anindefinite determiner a an

indefinite determiner

🔗 ðuðu ⟨ðu⟩ — a; an; thegeneric determiner, used to refer to a complement in a general, nonspecific sense a an the

generic determiner, used to refer to a complement in a general, nonspecific sense

🔗 caycay ⟨cay⟩ — change; evolve change evolve

🔗 cici ⟨ci⟩ — bymarks the complement as an agent by

marks the complement as an agent