The cmavo of selma’o NOIhA are all the hype right now — everyone‘s talking about them.

I wish.

Somehow most people have never even heard of NOIhA. A good while ago, in March 2016 to be exact, I wrote what was to this day the only written documentation of NOIhA out there. But… it’s in Lojban:

I think very few people have read that article. Shocking, I know!


Well, anyway.

Since I consider NOIhA absolutely necessary, the best way to raise awareness is to write about it in English. That is the purpose of the present article.

Let’s not waste any time and jump right in.

Lojban does not have adverbs, neither as a part of speech nor really in any other way.

Some have said that seltau (the left part of a tanru) are like adverbs. Yes, cafne klama will usually be interpreted as “go often” (lo nu klama cu cafne), but xunre zdadi’u is more often “a house that is red” (xunre gi’e zdadi’u), and of course there are, for any tanru, infinitely many other possible readings. Another problem is that there is only one place in the entire bridi where a seltau can be placed: right before the selbri. It would be very impractical if adverbs could only appear right before the selbri. But tanru were not designed for expressing adverbial meaning precisely, and they cannot be considered a solution.

Some people like to compare tenses, tags, even negation, to adverbs. I do not disagree with this, but here we have to note that none of those things are predicates. There are a limited number of convenient cmavo that express certain adverbial meanings, but there is no general way by which to generate any desired adverbial meaning from any brivla. For a language whose main expressive power comes from its predicates this is bad news, obviously, because it means that the vast majority of available predicate meanings are not accessible from within adverb land. This is not general enough for our purposes.

Reading the above you may have thought to yourself “Wait, but what about fi’o?”. The cmavo fi’o does indeed turn any selbri into a tag (preposition/sumtcita), but the problem is that the relationship between the created tag and the bridi in which it appears is entirely unspecified. The precision of a fi’o tag is comparable to that of a seltau. The fi’o tag is somehow vaguely related to the bridi, the seltau is somehow vaguely related to the tertau (the right part of a tanru). fi’o gets us slightly further than a tanru, though, because at least it can be moved around freely in the bridi, unlike a seltau, and it attaches to the bridi directly, rather than to the selbri.

And those are pretty much the only options that exist right now when you want to express adverbial meanings. Something is missing, and I think I have an idea what it is.

There is a NOIhA-shaped hole in the language.

What the heck is NOIhA, then?

NOIhA is a family of cmavo that express adverbial meaning.

One can understand a statement like “Susan is running quickly” (where “quickly” is the adverb) as meaning something along the lines of

[∃e] (e is Susan running) ∧ (e occurs quickly)

This shows that adverbs modify bridi, and they do this by making additional claims about them.

This is also what the NOIhA family does. However, we will find that more than one kind of NOIhA is needed to cover all possible adverbial meanings.

Important: All of the cmavo of selma’o NOIhA you will encounter in this article are meant to be placeholders. They should get replaced with shorter cmavo at some point.

Let’s start with —

‹poi’a› and ‹noi’a›

poi’a does exactly what we saw above: it modifies a bridi by adding an additional claim that is connected to the main claim by an ∧ (“and”). What the added claim is depends on the predicate following poi’a. In very simple terms, poi’a turns a selbri into an adverb, which can then be placed anywhere in the sentence.

do dansu poi’a melbi
“You dance beautifully.”

and also

poi’a melbi do dansu
Beautifully you dance.” (same meaning)

The meaning of this poi’a sentence can be paraphrased like this:

lo nu do dansu cu fasnu gi’e melbi
“The event of [you dance] occurs and is beautiful.”

You can think of poi’a as being a poi that attaches to bridi rather than sumti. Just like poi, it restricts the thing it attaches to; it restricts the bridi. We can see what this means when we add quantifiers:

mu verba cu sanga poi’a xajmi
“Five children are singing comically.”

Here, the the quantifier five says how many things are singing comically, but it does not say how many things are singing (in whatever way). There may be things that are singing beautifully, weirdly, etc, but those are different properties which the poi’a xajmi bridi does not deal with. This is how poi’a restricts the bridi, like poi.

A realistic example of the contrast:

lo gerku pu so’i roi plipe .i ku’i no lo gerku cu plipe poi’a drani
“The dogs jumped many times, but none of the dogs jumped correctly.” (perhaps someone is failing at training their dogs to jump through hoops)

Now, we know that poi has a non-restrictive counterpart, noi. Similarly, poi’a has a non-restrictive counterpart, noi’a.

mu verba cu sanga noi’a xajmi
“Five children are singing, and they’re singing comically.”

Here, noi’a does not restrict the bridi to comical singings, so mu here counts the children that do any sort of singing, and then says that all of them sing comically. In this version there are no non-comical singings possible. This may be hard to grasp at first. If so, move on and return to noi’a later once the rest makes sense.

At the end of this article there will be expansions for all the NOIhA, but for now let’s move on to —


Before I explain what soi’a does, let us appreciate what having poi’a (or really, NOIhA in general) means for the language. Lojban has a lot of cmavo that express tenses, but the cmavo can sometimes appear quite unrelated to their source brivla. We recognize pu (from purci), ca (from cabna) and ba (from balvi) easily enough because we use them all the time, but some people have some difficulty learning FAhA for example. The tense form of zunle is zu’a (not zu’e), and that of pritu is ri’u (and not pi’u). What I’m trying to say is, you have to learn the cmavo forms separately. Even if you already know the words for “left” (zunle) and “right” (pritu), the tense that means “to the left” (which is exactly the same meaning that zunle has, but applied to the whole bridi) and the tense that means “to the right” have a completely different form from those brivla. And this is the case for every single cmavo you learn. With a tool like poi’a, you can always use the brivla you already know and don’t have to learn extra cmavo. Moreover, when you want to express a tense that has no associated cmavo (because cmavo space is limited), you can still express it using the appropriate brivla, as poi’a opens up the world of tenses to the potentially infinite brivla space.

So, to show a few examples of poi’a doing the job of familiar tenses:

mi sipna poi’a purci
“I slept.” (compare to pu)

mi sipna poi’a cacra
“I sleep for one hour.” (compare to ze’a lo cacra)

mi sipna poi’a cnita be lo jubme
“I sleep under the table.” (compare to ni’a lo jubme)


There are, however, certain tags where poi’a would give the wrong meaning, making it inappropriate as a paraphrase:

mi ka’e sipna
“I can sleep.” “It is possible for me to sleep.”

mi sipna poi’a cumki
“I sleep and it’s possible.”

but the meaning of ka’e sipna is not that sleeping actually takes place. Only the potential is claimed. The poi’a paraphrase, however, does claim both: sleep taking place and the possibility of sleep. The effect would be even more catastrophic with ka’e nai, since mi sipna poi’a na cumki is self-contradictory: “I sleep and it’s not possible.”

This is why another NOIhA is needed, one that only places the bridi into its x1 but doesn’t automatically claim the bridi. This is precisely what soi’a does:

mi sipna soi’a cumki
“It is possible for me to sleep.”

This is exactly equivalent to the ka’e sipna version. Instead of lo nu sipna cu fasnu gi’e broda, the expansion with soi’a is just lo nu sipna cu broda. You use soi’a for anything where you don’t wish to claim the main bridi. Examples are cumki, cafne, jitfa, lakne, se pacna, xanri, and so on.

do zvafa’i lo ckiku soi’a se pacna
“Hopefully, you [will] find the keys.”

soi’a lakne lo tcima ba mabla
“The weather will likely be bad.”

Like poi’a and noi’a, soi’a is sensitive to scope. The following sentence pair demonstrates how the meaning changes depending on soi’a‘s relative position to a quantifier:

pa lo pendo be mi soi’a se pacna ba speni mi
“One of my friends, hopefully, will be my spouse.”

expands to

pa da poi me lo pendo be mi zo’u lo nu da ba speni mi cu se pacna
“One of my friends is such that I hope that they will be my spouse.” (identity known)


soi’a se pacna pa lo pendo be mi ba speni mi
“Hopefully, one of my friends will be my spouse.”

expands to

lo nu pa da poi me lo pendo be mi zo’u da ba speni mi cu se pacna
“I hope that one of my friends will be my spouse.” (identity unknown)

Next up is —


While soi is not a member of selma’o NOIhA, it is rather closely related, so mentioning it here won’t hurt. Let me start out by clarifying that we’re talking about so-called new-soi here, not the soi that is described in CLL (which most Lojbanists dislike and avoid entirely). New-soi started out as a generic bridi relative clause (replacing xoi) and an early proposal described it in a way that you might now recognize as being similar to poi’a (and some of the examples look a bit like soi’a).

When I came up with NOIhA, the different possible functions of a bridi relative clause (which soi was) could be split into separate cmavo, and soi itself became redundant. A more specific use for soi had to be found, and this is the result.

In short, soi is a non-restrictive bridi relative clause with top-scope.

You know how in English, you can attach a “which” clause to the end of a sentence to talk more about it? “I was late, which led to trouble”, “You never listen to me, which really annoys me.” — This “which” is expressed by soi.

But unlike “which”, a soi-clause can be moved around freely. It most commonly appears at the end of a sentence, however, partly because it’s usually used as an afterthought:

ie’o lo carba’e na muvdu soi mi kanpe lo nu ke’a ba nabmi
“Well, the propeller isn’t moving, which I expect will be a problem.”

As you can see, the soi-clause contains a ke’a, whose antecedent is the bridi to which the soi is attached. It doesn’t matter where the soi-clause is placed; it always has scope over everything else.

A few more examples before we move on.

su’o prenu, soi ke’a nunkusru, cu xusra lo du’u mi spogau la .lojban.
“Some people, and this is mean, say that I’m destroying Lojban.”

mi fliba lo se troci, soi ke’a mabla
“I failed the attempt, which sucks.”

ca lo cerni mi gunka poi’a cacra be li so’i, soi lo nu ma kau ni frili pu spaji mi se’u, lo ka favgau lo cfika
“In the morning I worked for many hours, and I was surprised by how easy it was, on developing my fiction.”

ra pu se darxi pa drata verba, soi ra .aidji lo ka jungau lo mamta ke’a
“She was hit by another child, which she intends to tell her mother.”

‹poi’o’a› and ‹noi’o’a›

The adverbials poi’a, noi’a and soi’a cover almost all types of adverbs — almost. There is one other kind of adverbial exemplified by phrases such as “She left the room angrily.”

The difference between “leave angrily” and “leave early” is that it’s not the event that is angry, it’s the person who is angry.

In Lojban this sort of meaning is standardly expressed using fau:

ra cliva lo kumfa fau lo nu vo’a fengu
“She left the room together with her being angry.” “She left the room angry.”

The meaning of fau can be understood as follows:

broda fau ko’a = fasnu fa lo nu broda joi ko’a


broda fau ko’a = lo nu broda cu fasnu gi’e poi’i ke’a joi ko’a fasnu

An adverb like “angrily” is a fau-type adverbial that is tied to the x1 of the containing bridi (note the vo’a). fau lo nu vo’a broda is therefore the general way of expressing this kind of adverbial meaning.

While it’s nice to know that we have this option, there are certain practical disadvantages to this pattern:

  1. vo’a always refers back to the x1 of the top-most bridi, which means we can’t use it in more deeply nested structures. Subscripts can be used, or other pronouns, but then the pattern loses generality, which is also impractical, because it puts a big mental burden on the speaker.
  2. It will often require a terminator to prevent following sumti from being gobbled up by the fau abstraction.
  3. It is rather verbose.
  4. There is no non-restrictive counterpart. (fau‘s restrictiveness is shown by the gi’e in the second expansion above.)

For these reasons, the two adverbials poi’o’a and noi’o’a were proposed. poi’o’a replaces the fau lo nu vo’a of the previous examples:

ra cliva lo kumfa poi’o’a fengu
“She left the room angrily.”

You might be thinking: “This is not really much shorter than the fau version, what’s the point?” Well, remember that the NOIhA are only temporary placeholders that are supposed to get replaced by shorter cmavo eventually. The poi’o’a mechanism is both shorter in syllables and simpler in terms of structure.

Let’s consider the scope of poi’o’a. What’s the difference between

mu da denpa poi’o’a xagji
“Five things waited hungry.”


poi’o’a xagji mu da denpa
“Hungry, five things waited.”


We know that fau has scope, but what’s happening in this case is that the built-in vo’a causes the meaning of poi’o’a to depend on what’s in the x1 of the higher bridi. In other words, poi’o’a creates a new predicate that is the conjunction of the main predicate (denpa) and the adverbial predicate (xagji) and this new predicate (denpa gi’e xagji) is claimed to hold of the x1 (mu da).

mu da [ckaji lo ka denpa gi’e xagji]

is the same as

[ckaji lo ka denpa gi’e ckaji kei] fa mu da

because predicates (in this case lo ka denpa gi’e xagji) always have lowest scope. mu da broda and broda fa mu da are the same.

(I should note that gi’e is a bit of a simplification here, because we don’t only mean that both things are true, but that they are true as part of the same event. They happen together, at the same time, and form a bigger whole.)

The conclusion is that poi’o’a and noi’o’a do not have scope. Instead, they create a new selbri from afar.

noi’o’a is the non-restrictive counterpart. The difference between it and poi’o’a is the same as that between noi’a and poi’a.

mu verba cu sanga poi’o’a gleki
“Five children are singing happily.” (the number of singing children is unknown. only the number of happily singing children is known)

mu verba cu sanga noi’o’a gleki
“Five children are singing, and they are doing so happily.” (the number of singing children is known to be five, and all of them sing happily)

We have now seen every member of selma’o NOIhA!

The Grammar of NOIhA

Now that we know what the NOIhA mean, we can discuss the most useful syntax for them. The truth is, there are at least three known workable options, and picking the right one is very important in the long-run.

Let me first mention the one thing that is given. NOIhA create terms (in the Lojban grammar sense):

term_1 <- sumti / … / NA_clause KU_clause free* / adverbial

However, the question is how exactly the NOIhA clause should be formed.

The three options are:

1) adverbial <- NOIhA_clause subsentence FEhU_elidible

2) adverbial <- NOIhA_clause selbri FEhU_elidible

3) FIhO <- &cmavo ( f i h o / p o i h o h a / … / p o i h a / … ) &post_word

Let’s talk about each option.

Option 1

1) adverbial <- NOIhA_clause subsentence FEhU_elidible

This means that an adverbial clause would be a NOIhA followed by a full bridi containing a ke’a and terminated by fe’u. The full grammar would be:

term_1 <- sumti / … / NA_clause KU_clause free* / adverbial

adverbial <- NOIhA_clause free* subsentence FEhU_elidible

NOIhA_clause <- NOIhA_pre NOIhA_post
NOIhA_pre <- pre_clause NOIhA spaces?
NOIhA_post <- post_clause

NOIhA = &cmavo ( p o i h o h a / n o i h o h a / p o i h a / n o i h a / s o i h a ) &post_word

Advantages Disadvantages
  • Gives a lot of control over the placement of the ke’a making it easy to express complicated adverbials
  • Often requires a terminator to prevent any following sumti from being part of the inner bridi
  • Requires an explicit ke’a to be precise.
  • More verbose.
  • Less optimal for simple adverbials.

This is the option that old-newsoi used (and xoi still uses). The soi described in the present article also works like this.

Option 2

2) adverbial <- NOIhA_clause selbri FEhU_elidible

Here, the adverbial would only contain a selbri, not a full bridi. Therefore, there would be no ke’a. The full grammar would be:

term_1 <- sumti / … / NA_clause KU_clause free* / adverbial

adverbial <- NOIhA_clause free* selbri FEhU_elidible

NOIhA_clause <- NOIhA_pre NOIhA_post
NOIhA_pre <- pre_clause NOIhA spaces?
NOIhA_post <- post_clause

NOIhA = &cmavo ( p o i h o h a / n o i h o h a / p o i h a / n o i h a / s o i h a ) &post_word

Advantages Disadvantages
  • Rarely requires a terminator (only right before another selbri)
  • Optimal for simple adverbials
  • Requires be to add additional sumti to the adverbial.
  • No ke’a means complicated adverbials are harder to express.

This is the option currently in use (the examples in the present article also use this grammar).

It needs to be said that even though this version does not come shipped with a ke’a, a ke’a can be obtained indirectly by using a poi’i-selbri as the adverbial selbri and using poi’i‘s ke’a instead:

mi zvasta poi’a poi’i mi pacna lo nu ke’a ba mentu li su’e mu
“I stay for what I hope to be no more than five minutes.”

poi’a poi’i may look ridiculous, but again, remember that poi’a is a placeholder to be replaced by something prettier and shorter, and note that there is a popular proposal to use voi for poi’i‘s meaning. Those things combined could make this a much more elegant construction. In any case, option 2 can achieve the same things as option 1, but tends to be shorter.

Option 3

3) FIhO <- &cmavo ( f i h o / p o i h o h a / … / p o i h a / … ) &post_word

This option simply moves NOIhA into FIhO. No new grammar is needed, and instead FIhO’s grammar is used as is. Since FIhO creates a tag rather than a term, it expects a sumti to follow it. One option would be to treat that tagged sumti as the first empty place of the FIhO’d selbri, like this:

do to’e lazni poi’a rinka so’i xamgu
“You are hard-working causing many good things.”

poi’a rinka tags so’i xamgu and puts it into rinka‘s x2. (This is of course not how fi’o proper works, but it would make sense for poi’a)

One problem with this syntax is that not every adverbial has additional places. Observe:

soi’a cafne ku mi bajra
“I run frequently.”

The ku is necessary to prevent mi from getting tagged (as in ca ku). Also, when this option gets mixed with the poi’i trick, it can get a bit confusing.

Advantages Disadvantages
  • No new grammar.
  • Certain simple cases work well.
  • Can often require a ku.
  • Not optimal for anything complicated.

All three options have advantages and disadvantages. Option 2 seems to be the most popular right now, though the be is a bit painful. Option 1 is nice but a bit more verbose. Option 3 uses existing grammar, but in a new way, and faces the restrictions of that old syntax.

We should make sure to analyze each option thoroughly and pick the best one.


We have seen that NOIhA lets us do the following things:

  1. Express tenses in terms of selbri (poi’a)
  2. Express CAhA and other subordinating tags in terms of selbri (soi’a)
  3. Create infinitely many new adverbs from any brivla in a completely regular way.

What we need now is for the community to work together on the following two points:

  1. Picking the right grammar for NOIhA. This is important because we don’t want to be stuck with an option that wastes syllables or forces us to speak in ways we don’t feel comfortable in.
  2. Finding appropriate short cmavo to replace the NOIhA with. The short cmavo space is more or less completely used up (CV are all gone, CV’V have very few left). Therefore it will be necessary to repurpose older cmavo and this has historically been difficult (a taboo, in fact). That’s why we need to look for and choose the Right Cmavo together. The Right Cmavo should form a coherent series, be pretty, and short. The gain is so enormous that it will be worth it, but it may take some effort to convince enough people and get them to help finding the Right Cmavo.

With both points, we should try to be quick. NOIhA are already being used by people, and the longer we wait, the harder it may become to make the switch to the new replacement cmavo (and possibly syntax).

So, tell your neighbo(u)rs about NOIhA.

If you have questions or comments, leave them below.

Thank you for your time.

Appendix: Expansion Tables

Below you can find several expansion tables, which describe the most important scope interactions:

ko’a broda poi’a brodo
poi’a brodo ko’a broda
lo nu ko’a broda cu fasnu gi’e brodo
ko’a broda noi’a brodo
noi’a brodo ko’a broda
lo nu ko’a broda cu fasnu .i lo go’i cu brodo
ko’a broda soi’a brodo
soi’a brodo ko’a broda
lo nu ko’a broda cu brodo
ko’a broda soi ke’a brodo
soi ke’a brodo se’u ko’a broda
ko’a broda .i lo nu go’i cu brodo
PA da broda poi’a brodo PA da zo’u lo nu da broda cu fasnu gi’e brodo
poi’a brodo PA da broda lo nu PA da broda cu fasnu gi’e brodo
PA da broda noi’a brodo PA da broda .i ro da poi go’i zo’u lo nu da broda cu brodo
noi’a brodo PA da broda lo nu PA da broda cu fasnu .i lo go’i cu brodo
PA da broda soi’a brodo PA da zo’u lo nu da broda cu brodo
soi’a brodo PA da broda lo nu PA da broda cu brodo
PA da broda soi ke’a brodo PA da broda .i lo nu go’i cu brodo
soi ke’a brodo se’u PA da broda PA da broda .i lo nu go’i cu brodo
na ku ko’a broda poi’a brodo na ku lo nu ko’a broda cu fasnu gi’e brodo
poi’a brodo na ku ko’a broda lo nu na ku ko’a broda cu fasnu gi’e brodo
na ku ko’a broda noi’a brodo na ku lo nu ko’a broda cu fasnu .i lo go’i cu brodo
noi’a brodo na ku ko’a broda lo nu na ku ko’a broda cu fasnu .i lo go’i cu brodo
na ku ko’a broda soi’a brodo na ku lo nu ko’a broda cu brodo
soi’a brodo na ku ko’a broda lo nu na ku ko’a broda cu brodo
na ku ko’a broda soi ki brodo na ku ko’a broda .i lo nu go’i cu brodo
soi ki brodo se’u na ku ko’a broda na ku ko’a broda .i lo nu go’i cu brodo
‹poi’o’a› and ‹noi’o’a›
ko’a broda poi’o’a brodo ko’a broda fau lo nu vo’a brodo
ko’a broda noi’o’a brodo ko’a broda .i lo go’i cu brodo
PA da broda poi’o’a broda
poi’o’a brodo PA da broda
PA da zo’u da broda fau lo nu da brodo
PA da broda noi’o’a broda
noi’o’a brodo PA da broda
PA da zo’u da broda .i ro da poi go’i zo’u da broda fau lo nu da brodo
Categories eng

4 thoughts on “NOIhA

    1. ki’e do .i mi ji’a pacna

      (to xu do pu skudji su’o simsa be «lu .a’o snada lo ka zatfa’i lo cmavo poi basti kei je lo ka ciksi lo cnino co’e li’u» .i mi na jimpe tu’a «lu noi’a prami li’u» toi)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s