Monday, November 24, 2014

Paiodd Discourse Continued



To continue my discussion of Paiodd discourse, I thought we could examine alternative word orders that are still grammatical. Again, the typical word order is OSV, which means that, generally, the information in the Object slot or complement is the new information. See example (1):

(1)   Rodd lodd az. 'The man sees a woman.'

The chief rule of Paiodd word order is that the S and V must be directly adjacent at all times. This gives the alternative word orders of VSO, OVS, or SVO. As mentioned in the previous post, VSO often occurs in sentence-final clauses. If, however, it were the only clause, the effect is to emphasize the action itself, as in (2):

(2)   Az lodd rodd. 'The man sees a woman.'

A context in which this might occur is shown in (3):

(3)   Fazrir lodd fuian arioé rodd şopía. Oz lodd rodd. Dur sem izío asía xíamir.
'A man was working in the fields when a beautiful woman walked by. The man saw the woman. He felt love rising in his heart.'

OVS and SVO both appear to emphasize the post-verbal element. This can be tricky, as one must rely on context to know which is actually the subject. Sentences (4) and (5) show these two word orders.

(4)   Rodd az lodd. 'A man sees the woman.'
(5)   Lodd az rodd. 'The man sees a woman.'

To provide contexts where (4) might be a plausible sentence, consider (6):

(6)   Fazilía rebir rodd ario. Serss şopía ó alía, ai peað ðaiyo, þunlugér yiap dir ó ðimir. Cemuí, rodd oz lodd.
'A woman was walking near a field. It was a beautiful day, (as) the birds were singing, and butterflies darting to and fro. Suddenly, a man saw the woman.

In (6), the man is highlighted as important new information by being placed at the end of the clause.
            In (7), however, it is the woman who is highlighted as new information, even though she is in fact the object of the sentence. This highlighting takes place by post-posing the woman to the end of the clause.

(7)   Fazrir lodd fuianía. Serss şopía ó alía, ai peað ðaiyo, þunlugér yiap dir ó ðimir. Cemuí, lodd oz rodd.
'A man was working in his fields. It was a beautiful day, (as) the birds were singing and the butterflies darting to and fro. Suddenly, the man saw a woman.'

Now, the same focus on rodd could be accomplished by using the default word order, but perhaps with a slight loss of emphasis. For example, (8)

(8)   Fazrir lodd fuianía. Serss şopía ó alía, ai peað ðaiyo, þunlugér yiap dir ó ðimir. Cemuí, rodd lodd oz.
'A man was working in his fields. It was a beautiful day, (as) the birds were singing and the butterflies darting to and fro. Suddenly, the man saw a woman.'

In this case, the woman is still new information, and even retains a high degree of importance, but the default word order is not as striking or out of the ordinary, thus rendering the woman slightly less emphatic. Perhaps the author will elaborate on her importance later, or perhaps the woman was not particularly noticeable in this case. For instance, perhaps the woman was the worker's mother or someone else that would not attract particular notice. On the other hand, post-posing the object to final position in (7) implies that the woman stood out in some way, likely by being especially attractive, and probably someone the worker does not know.

Promoting Focus

One use of the deictic suffix -en 'this' is to promote a just introduced focus to the topic of the next sentence. For example:


In contrast, the suffix -an 'that' can be used to bring an item back into focus, as in (10), or to make a statement about a general category, as in (11).

(10)                       [continuing from (8) and (9)] Sersan visenía siepé. 'That day was the greatest in his life.'

Sentence (10) is recalling the 'day' introduced in (8), in order to make a further comment on it.

(11)                       Ladan az şopía rodd, filan sem ía. 'That man who sees a beautiful woman knows desire.'

In (11), ladan 'that man' refers to any man who finds himself in the stated situation, and thus does not have a particular referent, but merely states a category.

Friday, November 14, 2014

Paiodd Discourse

Currently taking a Discourse Analysis class. It's got me thinking about how some of these things work in Paiodd. Normal word-order for Paiodd is OSV, though there is some wiggle room, so long as the S and V are directly adjacent (a weird quirk left-over from my pre-linguist days, when it seemed like the subject had a much more important connection to the verb than any objects; in other words, before I understood properly what a linguistic predicate was). Anyway, given that this is the standard, default word-order, it would suggest that generally, the comment, or new information is part of the predicate, the O, while the known information is the S, and perhaps the V as well. Yet I'm not sure that's necessarily the case, as I think it is possible to translate the following with indefinites:

(1)   Selor pivar weloun. 'A king commands a soldier.'

Actually, seeing it now, I don't think it does work. This sentence would normally be interpreted as 'The king commands a soldier', though 'The king commands the soldier' is also possible. So it would seem that my intuition is probably correct. New information CAN go into the O slot, though it is possible that it may be old information as well.

            If we consider the other logically possible orders, given the strict rule for SV adjacency, we have: VSO, OVS, and SVO. VSO is almost always the word-order for dependent clauses, as in (2):

(2)   Guír pivar oz eusé selor zir. 'The king saw his wife when the soldier entered the hall.'

In this case, the VSO clause is the second clause. It is possible to reverse the order, however, so that the dependent clause comes first, while the main clause is postposed to second position, as in (3):

(3)   Zir selor eusé oz pivar guír.

The meaning is roughly the same, though it sounds to me as though the wife is the soldier's, not the kings now. That's pretty interesting! To specify that it was the king's wife, I would simply use the possessive pronoun:

(4)   Zir selor eusé oz pivar íasem guír.


Well, I've just discovered something about Paiodd discourse, even though I can't entirely understand why that is. It does seem, however, that in the case of (3), the focus (comment, or new information) is on the wife, connecting her to the initial topic (known information) of selor 'the soldier'. In (2), however, both the king and his wife are the known entities (part of the topic), while the entry of the soldier is the comment/focus/new information, and there is an implication that the wife had a noticeable reaction to this new event. In (3), however, the implication is that the king was impressed by the soldier's wife.

This is an excellent example of how conlangs take on a life of their own, as well, since I did not intentionally think these things through, but I find them to be true in the language nonetheless. It certainly may be due to some influence from my native language or others that I'm familiar with, but it was not a conscious decision. I will no doubt continue to think through these things, and I hope to either update this post later or write another focusing on the use of other possible word orders and their effect on discourse.

Sunday, October 5, 2014

A New (ish) Conlang: Ittanyo

Spending the afternoon reworking the former Ittahoe, which, under the new rules, is now Ittanyo. Even got it to a good enough point to translate the Lord's Prayer. So here it is!

The Lord's Prayer

Ran madibanyo tadoga mo ri.
Tia pina saroba.
Padira ye padoga naginal ko ye ndaristal anocha talo.
Adia tarobanyo tat yeda tidat,
ye ngadia ndanochanyo tap tipta sa,
wo madia ndanochano tap mbalombe adia lanocha tap tapto tipta sa.
ye parolan tia tila sa,
tu pidabo tia talo.

You are our father in heaven.
We praise your name.
On earth and in heaven, let your kingdom come and your will be done.
Give us today our bread of the day,
and do not charge our bad deeds to us,
as we do not charge the bad deeds of those who incur debt against us against them,
and do not lead us into temptation,
but free us from evil.

The Old Version

The original version was based on two main ideas: 1. that every inflectional form would be accomplished by changes in syllable structure (retaining the same phonemes, but re-ordering them into different syllable structures without adding or taking away anything), and 2. That the main inflections would be for "valency" of both nouns and verbs. That is, verbs would rearrange their syllable structure depending on how many arguments they took, and nouns would rearrange theirs depending on how many verbs to which they acted as arguments. I kind of liked how this made relative clauses extremely easy, so that you could say something like: Man woman love see = The man sees the woman (whom) he loves. You could tell that 'man' was the subject of both verbs by its form, and 'woman' was the object of the same two verbs. It was a cool system, but I wanted to try a slightly different approach.

The New Version

In the new version, I took some inspiration from Bantu noun-class prefixes, and another (though similar) idea I had had before: to create a language where nouns in relative clauses would have different case-marking than those in main clauses. I used prefixes inspired by Bantu noun-classes to indicate this. It is somewhat similar to the previous version, in that the important feature of noun inflection is which verb or verbs it is an argument for, but prefixes, rather than rearranged syllable-structure indicate this now. A good example is the first sentence of the Lord's Prayer above.

Ran        ma-             diba  -nyo  ta-            doga    mo                  ri.
You       nom/nom-  father -our  obl.REL- heaven  be.located      are

nom/nom = nominative of both verbs.
obl.REL = oblique argument of the verb in the relative clause.

Note: I realized after posting that I forgot an entire line: Thy kingdom come, thy will be done, on earth as it is in heaven. I'll fix that and post the corrected version when finished.

Wednesday, August 13, 2014

Linguistic Feature of the Week!

In an effort to keep this blog at least moderately active, I'm going to try to at least post an interesting linguistic feature once a week. This week: Switch reference! Here's what wikipedia has to say about it: http://en.wikipedia.org/wiki/Switch_reference

Thursday, April 17, 2014

Applying Total Physical Response Techniques to Learning Your Own Conlang

       A popular (for good reason) technique in language teaching in recent decades has been what is known as the Total Physical Response (TPR) method. This method involves very hands-on, practical, game-like activities that force you to actually use the language you're learning. Although it shouldn't necessarily be used exclusively, it is extremely useful in the early stages of language learning. I'm currently taking a class on Second Language and Culture Acquisition (SLACA), in which we are using these techniques to learn a minority language, and already, after only two sessions, I can attest to its effectiveness.
As a conlanger, I've been wondering how we can apply some of these methods to learning our own conlangs. Although he is not the originator of TPR, I will base much of this post and subsequent ones on the ideas of Greg Thomson, some of whose work can be found at his blog: https://growingparticipatorapproach.wordpress.com/. Thomson identifies six phases of language (and culture) learning, a rough outline of which is available here: https://growingparticipatorapproach.wordpress.com/an-idealized-program-six-phases-of-super-charged-activities/. The initial stage is referred to as the “Here and Now” phase, in which the goal is to begin learning vocabulary that is entirely practical, and refers to things that are concrete, “here-and-now”. I will begin by describing how to apply TPR to these “here-and-now” words in a normal language learning situation, then describe some of the problems for that approach to learning your own conlang, and suggest some modifications that will help to counter some of these problems.
In our first language learning session, directed by our SLACA instructor, who explained to our language consultant what to do, the goal was to learn what Thomson calls a “dirty dozen”, that is, roughly a dozen to fifteen words which are concrete and practical. We were not allowed to repeat these words or write them down. Our job was to hear them, and be able to identify the correct items when called upon. At this point, we just want to hear the words, and learn their meanings, without analyzing anything more. That part comes later.
We learned basic actions, such as “sit”, “stand”, “walk”, and “run” by our language consultant demonstrating the actions. We began with just two, “sit”, and “stand”, which were repeated several times. Then we added a third, which forced us to choose and listen more carefully, and then finally the fourth word. We also learned “lie down”, so the exact number is somewhat flexible. The key is to begin with just two, with lots of repetition, the slowly add more words into the fray.
To learn words for basic objects, the first step was again to take two words, which were, in our case, “man” and “woman”. Our teacher elicited these words from our language consultant (a native speaker), by placing two dolls on the table, one a man, one a woman. The native speaker gave us the words for each, then asked us to indicate one or the other, depending on the word she said. We could point to them, but pointing may not be appropriate, depending on the culture, so instead, we were told to extend our open hands toward the appropriate item. After several repetitions, another item was added, a “boy”, this time. Now, we had to choose between three different items, man, woman, or boy. Again, after some repetition, we add a “girl” to the mix, and again we have to indicate the appropriate items based on what the native speaker says.
At this point, since we were in a small group, our language consultant began to call on us individually to indicate the various items, taking turns. Thus, we could now actually pick up the dolls representing each item. After doing this for a while, we began to add figurines of various animals, and the whole process was repeated. By the end, we had roughly a dozen words with which we were familiar. To be sure, we had not necessarily memorized these words, but we had enough familiarity with them that we could recall them the next session, two days later. This same format can be used for many other types of vocabulary, so long as it is concrete, “here-and-now” vocabulary. As a learner, you can take responsibility to set goals for yourself about what you want to learn. I'll post further ideas in the coming weeks, I hope, but this gives you an idea of how TPR activities work.

Problems for learning your conlang

However, there are several issues with applying this to learning your conlang. Perhaps the most glaring is that there are no native speakers to consult. Quite likely, you are the only one who speaks the language. This means that manipulating objects and saying the names for them is relatively easy, but how are you going to respond to someone else indicating objects or giving commands? Fortunately, modern technology gives us some options. By recording yourself saying the names of various objects or actions in random sequence, and listening to this recording at another time, you can still get the benefit of TPR activities for learning your own conlang. Here's what I suggest!

Suggestions to Compensate

First, you should plan a significant amount of time. In my SLACA class, our sessions are about 40 minutes to an hour. I suggest no less than 20 minutes, but probably at least half and hour. You'll have to plan accordingly, with two or three different activities or sets of a few words each to learn. For example, in the first session, you might want to have four to five actions, four or five words for people, and another four or five words for animals.
On your first day, or your first session, simply focus on saying the words for actions and objects  while doing or manipulating them. Break it up into sections, if you prefer, beginning by saying and doing the actions at the same time. In a 30-minute session, spend perhaps 5 minutes doing this. Then record yourself saying the actions in a random sequence. You can perform them as well while you are recording, which will further help you remember the words, but it isn't strictly necessary. The goal is to listen to this recording at your next session, and do the actions as you listen.
After you've recorded about five minutes of saying the actions in random order, move on to identifying objects, using the same method. That is, using dolls, action figures, or simple pictures of the people (or animals, or objects) in question, say the word for each item while picking it up or pointing to it. Do this for about five minutes again, then record yourself saying the words in random order. Again, it will be helpful if you continue to do the actions while you record, but the goal is to listen to the recording again at another time, and perform the actions then. Then, once again, move on by adding animals (or 4-5 more words of whatever your target vocabulary set is), and perform the procedure again.
By now, you should have 12-15 words to work with. Depending on how much time you allot yourself, this is probably a good place to stop. If you have more time, or want to take more time, you can combine the actions you have learned with the people and animals. So, for example, you might say in your conlang, “The horse is walking” (or “The horse walks”, to keep it simple if your conlang has progressive forms), and use the figurine or picture of the horse to make it walk. Then you might say, “The man is sitting”, and again make the figurine or picture of the man sit (if you use pictures instead of dolls or figurines, you might want to have a picture of a man, or woman, child, dog, etc, sitting, and others of him walking, running, and standing; you will then pick up or point to the appropriate picture as you say the sentence). Once again, after doing this five to ten minutes, record yourself saying appropriate sentences in a random order.
The next step is to give it a rest! You've been hard at work, and your brain needs time to process all this information! Take a break, relax. Then, the following day or a couple of days later, listen to your recordings and respond to the sentences appropriately – sit when you hear yourself say “sit”, make the figurine of the man stand when you hear yourself say “The man stands”, or simply point to or pick up the figurine or picture of the rooster when you hear yourself say the word for it.
After listening to the recordings and responding to them appropriately, you might wish to add some new words, using the same type of activities described above. Or perhaps you might spend an entire session reviewing the recordings to solidify what you learned the last time. You'll have to decide what works best for you. I intend to post a video demonstrating these techniques soon, and in the coming weeks and months, as I have time, I'll describe other activities that can help you learn your conlang in fun, interactive, and engaging ways. In addition, for other ideas, I'd encourage to check out Greg Thomson's blog linked above! Until then, happy conlanging!