INTRODUCTION: Why bother?
Language is an essential part of any culture. Culture wouldn't exist without some form of communication, and humans communicate by speaking. But the connection is deeper even than that. The language helps to define the culture which uses it, and is in turn affected by that culture. This dynamic process is what keeps a language 'alive': Latin is a dead language because the culture to which it belonged no longer exists.
In the creation of a campaign world, it is entirely possible to invent histories, people, places without any reference to the language except to create a few fantastically-sounding placenames like 'Edroleth', 'Calanzia' or 'Peredria'. But the people of this world will think just the same as you and I. Their language will not be sufficiently different to give them a different perspective on life, a different set of assumptions. Unless you are an exceptionally talented, original creative-writer your histories will become imitations of Earth's own: mass-produced and a little monotonous.
But is it worth the trouble of creating your own languages just to give your campaign worlds that extra bit of authenticity and personality? To quote an example: would the world of Tolkien's Silmarillion have been the same were it not for his invention of Quenya? Tolkien's work is indisputably amongst the most original fantasy ever written. And Tolkien was a philologist.
Even so, is it necessary to create a detailed grammar and lexicon for the language, when a few words and phrases and an esoteric-looking alphabet could suffice? I believe so, and it is not all so difficult if you have a good understanding of the structure of languages.
Near where I live, the placenames are primarily a mixture of Norse and Anglo-Saxon. An awful lot of history can be contained in a single placename. Imagine: a party stumbles across a valley with a rotting wooden sign written in Ancient Ugralian. 'Sagrod mu Ctagrolis' it reads. One of the party (probably a mage) can read this and knows a little bit of ancient Ugralian. 'Sagrod' means valley, he explains, and I know '-olis' is a suffix meaning 'big', but I've never come across 'Ctagar' before... The suspense is built: The valley of big what? It could mean 'big waterfalls', or 'big dragons' the PCs just don't know. They proceed with caution.
******************
INVENTING YOUR LANGUAGE
So where do you begin with your campaign world? Do you start with a language and build up the civilisation around that or vice versa? I find it best to do both simultaneously: this is probably most realistic also, since language and culture developed contemporaneously and influenced each other. You think what sort of culture you want, and you imagine how you want the language of that culture to sound. Think how the guttural Klingon language with its profusion of glottal stops and unpronounceable consonant groups reflects the violent Klingon society.
A relaxed society, on the other hand, would not have such percussive sounds, or would give them less dominance. Sounds like 'sh', 'm', 'l' would be preferred, and complex diphthongs would result (like the French 'aille' or the Gaelic 'aou'). There are more likely to be 'silent' (i.e. non-pronounced) letters in such a language whereas in a more rigorous society all letters are pronounced (like the final '-e' in German words which in English would simply produce a mutation of preceding vowels).
Once you have decided upon a general mood for your language, and decided on the most prominent vowels and consonants, I find it helpful to talk nonsense using these sounds. You'll find certain combinations you like which you can then use as words or phonemes (bits of words). Try writing down the sounds you make and developing a standard way of doing this: e.g. 'shoneesh esculpad sad warisna hi catitow'. The next stage is the development of an alphabet.
The currently accepted theory of how alphabets developed is that their letters are abstracted forms of pictures. Hieroglyphs are a little unwieldy to continually copy out, so a shorthand is simpler. If you now take a picture of an 'owl' to represent the sound you make when you say 'owl' (i.e. an 'ow' sound) then abstract that picture into a letter you would get a letter representing the sound 'ow' (actually 'ow' is a diphthong so most languages would represent it as a juxtaposition of other letters). For instance, the cuneiform alphabet took off around 3000BC, developed from earlier Sumerian picture-languages.
So imagine a few common nouns associated with particular sounds e.g. bird might be 'tschismi' from an onomatopoeic association - that's what birds sound like (OK some birds sound like that, not many, but a few). The predominant sound in 'tschismi' is 'tsch' so a letter could be invented to represent the sound 'tsch', which looked like a scrappy drawing of a bird. If you try drawing a simplified bird a few times until you can draw it fluently and quickly, the resulting 'picture' can be adopted by your alphabet.
—| /
. |/
. |
(pardon my ASCII)
An alternative method is to simply scribble something suitably elvish-looking on the back of an envelope and isolate individual letters from that. That way gives you more control over the aesthetics of your alphabet, and a mixture of the two approaches is probably your best bet.
Now it's time to consider placenames. If you have already developed your campaign world, it's best not to change any of the names in case you end up with inconsistencies, so you should examine the names and break them down into individual words and ask yourself what these words might mean. Take a fictional town called 'Irsolan' which is built by a lakeside in the middle of a forest. 'Irso' could then be taken to mean 'lake' and 'lan' to mean 'glade'. Or alternatively 'Irso' could denote 'beside' as in the Scottish 'Kyle' and 'Lan' would then denote 'lake'. You then have your first few words.
A lexicon can now be compiled by continuing this method or simply by inventing words that sound good.
Moving on to grammatical considerations...I will take the time to give a cursory account of basic grammar for those unfamiliar with ideas of inflexion (conjugation and declension) or syntax.
******************
In most western languages words change according to their usage. In English for instance 'I tell' becomes 'he tells': an '-s' is appended to the root 'tell' when we're talking about someone else telling. In the case of verbs (words of action) this change is known as conjugation. If we list all the possible different people who could be 'telling' we find we can group them into six main categories (more for some languages):
I tell (first person)
you tell (second person)
he/she/it tells (third person)
we tell (first person plural)
you (plural) tell (second person plural)
they tell (third person plural)
Obviously English has lost most of these changes and only in the 'he/she/it' (third person singular) case is any change apparent. In German, though this distinction is much more exaggerated:
ich spiele
du spielst
er spielt
wir spielen
ihr spielt (or Sie spielen)
sie spielen
When we are talking about the past, things get more complicated. In most languages there is a simple conjugation we can do to shift to the past tense (e.g. English 'I played', an -ed is appended; Latin 'amabam' an -abam is appended). This is called the simple past or imperfect tense. There is often a 'perfect' tense construction: 'I have played', 'ich habe gespielt'. Here we have something more complex than a simple conjugation. The verb is converted into a 'past participle': played, gespielt, and an 'auxiliary verb' (here 'I have', 'ich habe') is used. This is something of a blessing: it means we don't have to learn another set of conjugation endings for the perfect tense (whereas in Latin we would: 'amavi' is the perfect of 'amare').
There are other tenses: future and pluperfect for instance. They all have their uses. There are more complicated ideas such as passive voice, subjunctive mood, etc. which find their way into most languages. A fuller account of such tenses is given at the end of this essay as an aid to GMs who want to create their own complex languages, but their explanation is unnecessary here.
Nouns (object words) also change according to their use in a sentence. In English, 'he' becomes 'him', 'bird' becomes 'birds'. In a sentence there is always a verb, and the person who is doing that verb is called the subject. For instance: 'I clobbered the orc'. 'I' is the subject here because 'I' is doing the clobbering (NB: 'I' is not trying to sound like Ali G here). The orc is then called the object. It is having the verb done to it. Many languages make a distinction between these two cases: in the first case (the nominative or subject case) the noun takes on one form, in the second (the accusative or object case) the noun takes on a different form. 'He hit me', 'I hit him': 'he' is the nominative, whereas 'him' is the accusative, and similarly with 'I' and 'me'.
The changing of nouns to indicate their status in a sentence is called declension, and there are often a few more cases than just nominative and accusative: the genitive case indicates possession (e.g. 'his', 'the bird's') and the dative case indicates an indirect object. An indirect object is like 'to him' in the sentence 'I gave it to him'. Here 'I' is the subject. I gave 'it' to him, so 'it' is the object. 'him' is called the indirect object, the recipient of the object. Basically a dative case is used wherever a preposition is used (a word like 'to', 'from', 'with', 'in'). In some complex languages like Latin and Greek more cases still are identified (the ablative and the instrumental) which are basically just variations of the dative for particular pronouns.
The other example of declension is to denote a plural: in English this consists of adding an '-s' to the end of a word, in other languages it varies. But for each case we have so far identified (nominative, accusative,...) there is a corresponding plural case (nominative plural, accusative plural,...).
This whole system of conjugation and declension is called inflexion, and is a necessary characteristic of conventional languages. Some languages are more inflected than others: German and several Eastern European languages are rigorously inflected, as are most ancient languages like Latin, Greek and Sanskrit. French and English are less rigorous and only really reserve distinctions for tense, third person, plural and genitive.
Aside from inflexion, the syntax of a language is also important. By syntax I mean whereabouts in a sentence a word goes. In a simple German sentence: 'Ich spiele Fussball' the verb (spiele) is the second entity in the sentence, preceded by the subject and succeeded by the object. The indirect object is more difficult: when it's just a pronoun (I, you him...) it would go before the object (e.g. 'Ich gab ihm den Fussball' - I gave him the football). When the object is a pronoun, it would precede the indirect object: 'Ich gab ihn dem Fussballspieler' (I gave it (the ball) to the footballer). In English the same is generally true. Subsequent verbs in a German sentence are placed in reverse order at the end of the sentence in their 'infinitive' form (the form you'd find in a dictionary). In English it's quite different, because the verbs are placed straight after the first verb in the left-to-right order in their infinitive form. (NB usually the Germans find a way around ungainly verb constructions at the end of the sentence: 'I want to continue to clobber him' becomes 'Ich will ihn weiter schlagen', with 'will' as the first verb and 'schlagen' as the only verb at the end of the sentence - 'weiter' means 'more' or 'further').
In Latin, however, the use of pronouns as subjects is not common (the conjugation of the verb is relied on, so 'amo' means 'I love' while 'amat' means 'he loves') and the sentence structure is entirely different. The verb tends to go at the end of a sentence and the declensions are relied upon much more heavily to understand the words' relationships to one another.
******************
So when it comes to the next stage, inventing a grammar for your language, where do you begin? First, consider how complex your language should be. Do you really want fifteen different ways of declining nouns? How many cases do you need to have?
Consider how advanced your civilisation is. Primitive civilisations have primitive languages, which generally become more convoluted as clans interact and complex grammars are formed. Eventually a language like Latin or Greek evolves with lots of declensions, conjugations, tenses, moods and voices. But does a complex society demand a complex language? For instance, English is relatively simple in its inflexions and present tense conjugations, with its spelling irregularities in the perfect and imperfect tenses arising from its Saxon origins. And yet English is used by societies far more complex and diverse than Roman or Greek civilisations. The trend seems to be away from the multiplicity of conjugations and cases. For examples of this linguistic 'decadence' look no further than George Orwell's '1984' where Newspeak is the degenerate descendent of Oldspeak, and is considerably simpler and structurally more logical and concise. And is Orwell's language of compounds and simple constructions so dissimilar from today's English? 'email', 'cellphone', 'so 1980s'.
Anglo-Saxon is a fairly good-sounding language for fantasy writers, and it only has two tenses: the present and preterite (past) and ideas like future or conditional have to be gleaned from the context rather than the grammar. However it has a large number of cases for nouns (including an instrumental) and has more than the usual number of pronouns (ever wondered where 'ye' came from? It's derived from the second person dual pronoun in Anglo-Saxon).
Next, trawl through your set of phonemes (bits of words) which resulted from your gibberish earlier on. Pick out those that would sound good at the end of words, and use these as the endings for conjugated verbs or declined nouns. Set out tables for conjugation like I did earlier and similar ones for declension:
Singular:
Nom. shol
Acc. sholeb
Gen. sholeba
Dat. sholest
Plural:
Nom. sholi
Acc. sholebi
Gen. sholiba
Dat. sholesta
Deciding on how to express your ideas with a syntax, and using these tables you can now start to construct simple sentences, and move on to more complex constructions as you get used to the language.
******************
COMPLEX CONSTRUCTIONS
1. The Future:
English: I will do
German: ich werde...machen
Latin: faciam
Anglo-Saxon: ic fremme
Auxiliary verb constructions in English and German (auxiliary is 'will'), simple verb conjugation in Latin, present tense in Anglo-Saxon, with meaning of future implied.
2. The Subjunctive
This is not so much a tense as a 'mood'. It gives verbs an air of uncertainty and hypothetical usage: 'If I were to hit him...', 'H:atte ich ihm das Buch gegeben...' (had I have given him the book). In the same way as a noun has a set of singular cases and a set of plural cases, verbs have a set of ordinary (indicative) tenses and a set of subjunctive tenses. Its usage varies from language to language. In German reported speech (i.e. 'he said that we should all wear monocles') is expressed using the subjunctive. The subjunctive is often very similar to the indicative, but there are exceptions (like Latin).
3. The Conditional
English: I would do
German: ich w:urde...machen
German (imp. subj.): ich machte (also 'I would do')
Anglo-Saxon: ich wolde fremman
Auxiliary verb constructions in English and German (auxiliary is 'would', i.e. the imperfect subjunctive of 'will'). German also uses the imperfect subjunctive for this purpose. Anglo-Saxon uses the preterite subjunctive of 'willan', which is the precursor of the English and German construction.
4. The Pluperfect
English: I had done
German: ich hatte gemacht
Latin: feceram
Anglo-Saxon: ic fremede
Auxiliary verb constructions in English and German, but with the auxiliary in the imperfect rather than the present as for the perfect. As usual a simple conjugation in Latin (the vowel changes earlier in the word are simply an irregularity of the verb 'facere'. Anglo-Saxon uses the preterite to express the pluperfect.
5. Continuous past/present
This is essentially a feature of English which confuses people learning English and English people learning other languages, so it's best to point it out here. 'I was doing' and 'I am doing' are examples of this feature. It expresses an action which is done over a length of time, a process which is still going on. Most languages do not make this distinction: an imperfect tense replaces the 'I was doing' and a present tense replaces the 'I am doing'.
6. Alternatives...
There are some languages whose grammar is utterly different to the Latinised western tongues (e.g. Gaidhlig (Gaelic) the native language of Scotland and Ireland). 'Periphrastic' constructions (i.e. convoluted ways of saying things) are common. For instance 'Tha an taigh aig Domhnall' means 'Donald has the house.' Here 'tha' is the verb 'to be' (i.e. 'is'), 'an taigh' is the house and aig means 'at'. So we have 'Is the house at Donald', or (rearranged) the house is at Donald. You may want to experiment with long-winded or esoteric constructions and see where it leads you.
You must also consider whether you want to base your language on the conventional languages of Earth. How would a language be without any verbs? How would it feel to think in that language? Since language and culture are inextricably linked would it ever be possible to create a race of aliens if their language was structured like a terrestrial language? Would you not end up (metaphorically) with a Star-Trek situation where all races are humanoid because their languages are structured the same?
So I urge you: experiment with languages. Not only can it give your campaign world authenticity and an extra layer of detail, it can also help you think in different ways and create completely original worlds.