On 17 March 2014 03:24, Roman Rausch <[log in to unmask]> wrote:

> There is a nice book which is a treasure to all students of Japanese in
> particular, but could also be of interest to conlangers in general. It is
> called 'Handbook of Japanese Compound Verbs' (Yoshiko Tagashira and Jean
> Hoff; Hokuseido Press).

Sounds like a great book! COuld be a good addition to the LCS lending

> Compound verbs are just that, compounds of two verbs, but for some reason
> don't really exist in Indo-European.

Well, not in Western Indo-European languages, but they are common in Hindi
and other Indo-Aryan languages. They are also found in Modern Greek, but
are somewhat restricted in their form and use.

> In Japanese they range from literal (_tataki-kowasu_ 'hit-break = tear
> down, shatter') to abstracted (_yomi-ageru_ 'read-raise = 'read aloud'),
> idiomatic (_oshi-kakeru_ 'push-hang = intrude, visit without invitation')
> and culturally peculiar (_tati-yomi_ 'stand-read = read in a bookstore
> without buying', only in nominalized form).

That last one would make a nice addition to French, where that phenomenon
is common as well :).

> As you can see, they offer a great wealth of semantic finesse and
> scrumptiousness, potentially helpful and inspiring for one's own language.
Indeed :). And because of this inspiration, my own Moten has compound verbs
as well. They are really fun to work with. Two somewhat idiomatic examples
are _jaguba|si_, literally "to come and go", which can mean "to come and
go" (indeed!), "to move back and forth" and "to be nervous", and
_ipenlabutaj_, literally: "to sit and wait for", which can have this
meaning (referring for instance to waiting in a doctor's waiting room) but
generally means "to be patient with". Plenty of compounds with _istu|l_:
"to call, to summon" as the second (light) element have a verb as first
element as well, like _ipe|lastu|l_: "to show" (literally "to summon and
see") or _ipenlastu|l_: "to invite" (literally "to summon and wait").

> The mentioned book is a dictionary of 200 selected compound verbs, but it
> is also a paragon of presenting such material. For example, the first entry
> is _abakidasu_ 'dig up some criminal or confidential matter and expose it
> to the world'.
> The dictionary begins with giving the constituents:
> V1 _abaku_ 'expose some criminal or confidential matter'

That is one heck of a specific meaning for such a short stem! My dictionary
says it means "to disclose, to divulge, to expose" in general, which feels
more like it :).

> V2 _dasu_ 'take out, reveal'
> It classifies the compounds into types, this one is 'means-goal'.
> Then it comments on restrictions of usage:
> - a digging-up process has to be involved
> - there is an adverse effect on the target person
> - not used for physical objects
> - the exposure is public
> Finally, it lists several example sentences with the word, one being:
> _Imasara kako no koto wo abakidasatte shikata ga nai._
> 'It's no use raking up the past at this late date.'
> But also at least one incorrect example, if it is helpful:
> _**Omae ga kako ni mafia to kankei ga atta koto wo masukomi ni abakidasu
> zo!_
> 'We'll tell the media that you had connections with the Mafia in the past!'
> This is not correct because no digging is involved, one has to use
> _barasu_.

> Inspired by this book, I've been working on a similar dictionary for a
> similar semantically complex construct in my language Talmit (
>, which I call 'phrasal compounds'. So I
> hereby present to you the Talmit dictionary of phrasal compounds:
> Such compounds consist out of two monosyllabic lexical items joined by a
> grammatical element. The grammatical element can be a case postposition:
> - accusative _nu_: _tánugi_ 'thing-ACC-see' = 'to inspect, check whether
> sth. or so. is really there', _pánugi_ 'state-ACC-see' = 'to inspect, check
> the state of affairs'
> - genitive _mo_: _gámmoka_ 'nose-GEN-flat.area' = 'area between the nose
> and the upper lip'
> - dative _ma_: _dématal_ 'deed-DAT-word' = 'rule, law'
> - instrumental _za_: _mázagi_ 'eye-INSTR-see' = 'obvious, apparent;
> eyewitness'
> - themative _la_: _tállado_ 'word-concerning-ask' = 'ask for permission or
> to loan something'
> - nominative _ja_: _páhjamut_ 'shoe-NOM-nothing' = 'barefoot'
> It can also be the conjunctional suffix _-ra_ 'and' or _-se_ 'and,
> interacting with':
> - _básseba_ 'leg-with-leg' = 'gait; conduct, behaviour', _gíssepun_
> 'eye-with-finger' = 'skill', _áxeat_ 'one-and-one' = 'one by one, one at a
> time'
> - _dáhraba_ 'stone-and-leg' = 'measurement' (weight and length), _bórabno_
> 'mouth-and-mouth' = 'uproar'
> Finally, it can be _-ru_ which is adverbial, or participial after verbs.
> If the verb is repeated, this has iterative or intensifying meaning:
> - _árupsar_ 'good-ADV-rub' = 'warm up by friction', _írugi_ 'bad-ADV-look'
> = 'evil spell, curse'
> - _gíruga_ 'see-PART-go' = 'scout, spy', _dérode_ 'do-PART-do' = 'drive,
> thirst of action'
> As one can see, verbs joined like that are basically verbal compounds. But
> the second lexical item does not have to be a verb:
> - _gírumui_ 'see-PART-NEG' = 'invisible', _kézrumui_ 'pierce-PART-NEG' =
> 'invincible, not able to be pierced; inconceivable, difficult to get a
> grasp of', _írutru_ 'bad-ADV-inside' = 'non-edible'
That last one is a great idiomatic compound! I might just steal it for
Moten! :P

> Anyway, these are just some examples, there are currently 137 words in the
> list. In a perfect world world I would also add positive, and, where
> appropriate, negative example sentences to all the entries, but as of now
> you'll find that part rather sparse. It will hopefully grow with time.
> The entries are cross-linked with the etymological dictionary (
> The etymological dictionary has received an improvement as well, it now
> has a search mask which filters the table, and a clickable list of roots.
> Sound-symbolic roots which yield lexical items are also cross-linked.
> Comments are appreciated. Also, stay tuned for further updates.

It's indeed a great way to present compounds, especially given how they
work in Talmit. Are those the only kinds of compounds you have in the
language, or can you just put stems together without a joining element?
Christophe Grandsire-Koevoets
President of the Language Creation Society (

Personal Website: