Béu : Chapter 2 : The Noun: Difference between revisions

From FrathWiki
Jump to navigationJump to search
 
(35 intermediate revisions by the same user not shown)
Line 35: Line 35:
..
..


5) '''manga''' = a verb ... its base form (citation form) is equivalent to the infinitive in English (or some kind of verbal noun or gerund). When used as a verb, it will take its '''r'''-form, '''u'''-form or '''i'''-form.
5) '''manga''' = a verb in its base form (citation form). When used "actively" it will take its '''r'''-form, '''u'''-form or '''i'''-form.


An example is '''twá''' ... "to meet" or "a meeting"  (the concept of "meet" disassociated from any arguments, tense, aspect or whatever).  
An example is '''twá''' meaning  "to meet" or "a meeting"  (the concept of "meet" disassociated from any arguments, tense, aspect or whatever).  


..
..
Line 239: Line 239:
All prepositions that are not '''pila?o''' lead to the location being shunted into a relative clause. For example ...
All prepositions that are not '''pila?o''' lead to the location being shunted into a relative clause. For example ...


'''polga?o nài r bain gwai.a''' = "the sailing boat which is among the islands" or simply "the sailing boat among the islands"  
'''polga?o nài r fiagan gwai''' = "the sailing boat which is among the islands" or simply "the sailing boat among the islands"  


..
..
Line 246: Line 246:


'''bàu gèu dá''' = where is the green man ?
'''bàu gèu dá''' = where is the green man ?
..
Addendum ... To keep things simple I refrained from mentioning this before but time can also be expressed in this slot. For example "the turmoil '''jé''' 1300's". Actually location AND time can both be in this slot at the same time. For example "the anger '''pí''' Iceland '''jé''' 1950's ...".


..
..
Line 369: Line 373:
..
..


== ... Kenzuma==
== ..... Kenzuma==


..
..
Line 430: Line 434:
{|  
{|  
|-
|-
! gwai.a || nài  || polg-ai-r-a || ala || ʃì
! gwai.a || nài  || polg-ai-r-a || fía || ?ode
|-
|-
| the islands ||  {{small|REL}}  || sail-{{small|1PL.INC-IND-PRES}}  || between || them
| the islands ||  {{small|REL}}  || sail-{{small|1PL.INC-IND-PRES}}  || between || them
|} => "the islands that we are sailing between"
|} => "the islands that we are sailing between"


NP = '''gwaia''' : RC = '''nài polgaira ala ʃì''' => PC = '''polgaira ala gwaia''' ......................  '''nài''' is again binned. Also '''ʃì''' is discarded. The NP must be positioned behind '''ala''', the preposition that governs it.
NP = '''gwaia''' : RC = '''nài polgaira fía ?ode''' => PC = '''polgaira fía gwaia''' ......................  '''nài''' is again binned. Also '''?ode''' is discarded. The NP must be positioned behind '''fía''', the preposition that governs it.


In the above PC '''gwaia''' is not absolutive, also not adorned by a '''pila?o'''. Instead it exists in a prepositional phrase. For this reason, a pronoun '''ʃì''' is needed in the RC to represent the NP
In the above PC '''gwaia''' is not absolutive, also not adorned by a '''pila?o'''. Instead it exists in a prepositional phrase. For this reason, a pronoun '''?ode''' is needed in the RC to represent the NP


..
..
Line 494: Line 498:
Pronouns are used in partitives in the same way as NP's.
Pronouns are used in partitives in the same way as NP's.


'''a?a wì magi''' = "one of us"
'''a?a wì pài''' = "one of us"


'''ima wì ''' = "two of them"
'''ima wì onde''' = "two of them"


'''há ima wì glá timpura''' = "two of them are hitting the woman"
'''há ima wì onde glá timpura''' = "two of them are hitting the woman"


[ I guess English is a bit stange w.r.t. plural pronoun. I mean ... why not "*two of they are walking" {following the pattern "they are walking"} instead of "two of them are walking" ]
[ I guess English is a bit stange w.r.t. plural pronoun. I mean ... why not "*two of they are walking" {following the pattern "they are walking"} instead of "two of them are walking" ]
Line 769: Line 773:
..
..


== ..... Specifying the roll of a noun==
== ..... Pila?o==


..
..
Line 873: Line 877:
'''jenes solbori moʒi lé ʔazwo''' = "Jane drank water and milk"
'''jenes solbori moʒi lé ʔazwo''' = "Jane drank water and milk"


'''jonos jenes mwuri sadu ima lé ʔusfa uya''' = John and Jane saw two elephants and three giraffes.
'''jonos jenes hecuri sadu ima lé ʔusfa uya''' = John and Jane saw two elephants and three giraffes.


This word is that is never written out in full but has its own symbol. See below ...
This word is that is never written out in full but has its own symbol. See below ...
Line 879: Line 883:
..
..


== ..... Manga==
== ..... Maŋga==
 
..
 
These are verbs. In particular '''maŋga''' = "the infinitive form of the verb" or "verbal noun". They call it "maSdar" in Arabic.
 
I shall call it "base verb".
 
In the BLT ('''béu''' linguistic tradition) the base verb is considered "dead" or "inactive". Only when a suitable "tail" is added does it become "alive" or "active".
 
So "finite verb" is called '''maŋga ?algu'''. The infinitive form of the verb is simply  '''maŋga'''
 
(By the way ...  '''?algu''' = alive) ... the next Chapter is all about  '''maŋga ?algu'''.


..
..


These are verbal nouns or infinitives.
In '''béu''' the base verb is a noun. This is different to most languages in the world. In most languages, verbal nouns are derived from verbs. For example ... destroy => destruction : verb => noun whereas in '''béu'''. For example ... '''twá''' : "a/the meeting" => '''twarua''' : "I intend to meet".


English is very chaotic as to the various means it derive nouns from verbs. For example ... "discover" + "y" => the discovery ... "destroy" + "tion" => the destruction ... "run" + ∅ => the run. '''béu''' is as orderly as it is possible to get ... the verbal noun is in fact the base form of the verb.
..


[ Note to self : I need to do a lot of translation excercises to get my head around '''hulka''' = destruction = to destroy = destroying ]
English is very chaotic as to the various means it derives nouns from verbs. For example ... "discover" + "y" => the discovery ... "destroy" + "tion" => the destruction ... "run" + '''∅''' => a/the run. Whereas '''béu''' is as orderly as it is possible to get.


In this section I will usually translate a '''manga''' into its infinitive equivalent in English. For example ...
OK ... lets get started ...


'''solbe''' => "to drink"
'''solbe''' = "to drink"


Now the '''manga''' can amalgamate with other elements. For example ...
Now the '''manga''' can amalgamate with other elements. For example ...
Line 899: Line 915:
...and adding more elements ...
...and adding more elements ...


'''solbe moze sacois''' = "to drink the water quickly" or "drinking the water quickly"
'''solbe moze sacois*''' = "to drink the water quickly" or "drinking the water quickly"


The S or O argument in an active clause ... in the corresponding MP, must immediately follows the '''manga'''. Also because '''saco''' no longer immediately follows the  '''manga''', it must be explicitly tagged as an adverb by the -'''is''' suffix.
'''solbe moze sacois''' and '''solbe saco''' are examples of '''maŋguma''' ('''maŋga''' plus the augmentive  -'''uma''')


... and adding even more elements ...
Actually a  '''maŋguma''' can get as involved as a clause can. With arguments to the left of the head considered definite and arguments to the right, indefinite. But there is one difference. The ergative argument is marked by the particle '''hí''' rather than '''há''' or -'''ta''' or -'''s'''.


'''solbe moze sacois hí jono''' = "John drinking the water quickly" or "for John to drink the water quickly".
so adding even more elements ...


The A argument in an active clause,  ... in the corresponding MP ... comes last and has the particle '''hí''' in front of it. (the particle '''hí''' is probably related to the egative particle '''há''')
'''hí jono solbe moze sacois hí jono''' = "John drinking the water quickly" or "for John to drink the water quickly".


Note ... other clausal elements ( dative object, time, adverb, instrument, reason, purpose) can be added ... in our example they would come between '''sacois''' and '''hí'''.
This is as involved as I will go. But note that other clausal elements ( dative object, time, adverb, instrument, reason, purpose) can be added. As in normal clauses there is a tendancy to add them to the RHS.  


Now all elements from the active clause, that make it into a MP ... I refer to as "the manga heart".  
Now the '''maŋguma''' is basically a noun ... as is '''kenʒi''' and '''olus'''. And as with '''kenʒi''' and '''olus''' more modifiers can be added.


..
..


This manga heart can in turn amalgamate with other elements. These elements are pretty much the same ones that you find in a SP. The important thing is that the head has been replaced with a heart (my terminology). As you can see the heart has its own structure (as seen in the red box, below) ...
[[Image:SW_095.png]]I don't know the '''béu''' name for this. Actually I don't know the English name for this.


[[Image:TW_950.png]]
The items in the determiner slot are quite rare compared to  '''kenʒi''', but they are licit. There can be nothing in the quantity slot or the ownership slot.


As you can see above, the manga heart can take a slightly different form. The adverb can be put immediately after the '''manga'''. In that position it looses the -'''is''' suffix.
The "number + ''''''" modifier is quite rare. It means "so many iterations of the action". For example ...


The total amagamation ...  I call a MP "'''manga''' phrase". (As opposed to a SP "'''senko''' phrase" or a OP "'''olus''' phrase" ... all 3 being sub-types of noun phrase "NP")
'''uya wì hí pà solbe moze''' = (the) three times I drank water


You can see that a MP is pretty much the same as a SP, but with the '''senko''' (the head) replaced with a '''manga''' heart.
..
 
Adjectives can be a bit confusing. For example '''saco''' "quick" would be used immediately after '''maŋga'''. But if not immediately after '''maŋga''' but within the '''maŋguma''', it should take the form '''sacois''' (-'''is''' being the adjective => adverb suffix). But is it is outwith the '''maŋguma''' in the quality slot it will be '''saco'''. Got it ?


..
..
Line 929: Line 947:
In the example we are using '''sacois''' "quickly" can be taken out of the heart, and placed in the '''senko''' phrase as '''saco'''. In the adjective slot of course.
In the example we are using '''sacois''' "quickly" can be taken out of the heart, and placed in the '''senko''' phrase as '''saco'''. In the adjective slot of course.


Also you have a choice as to where you can place any locative. A locative can be placed in the locative slot of the '''senko''' phrase, or they can be placed in the heart, just before ''''''. For example ...
In a similar vain you have a choice as to where to put a locative. A locative can be placed in the locative slot or it can be placed in '''maŋguma'''. No change of form for the locative ...


'''solbe moze sacois tiapi hí jono''' = '''solbe moze sacois hí jono tiapi''' = "John drinking the water quickly in the house" = "for John to drink the water quickly in the house".
'''solbe moze sacois tiapi hí jono''' = '''solbe moze sacois hí jono tiapi''' = "John drinking the water quickly in the house" = "for John to drink the water quickly in the house".
OK ... so much for  '''maŋguma'''.


..
..


Now we have already introduced the '''pila?o'''. ''Theoretically'' all '''pila?o''' can be appended to a MP ... but most such constructions don't make much sense ... however  -'''tú''' appends often. For example ...
Now we have already introduced the '''pila?o'''. The '''pila?o''' are totally compatible with '''kenʒi'''. Nearly totally compatible with '''olus'''. However only two '''pila?o''' fit in with '''maŋga'''. These are '''pila?o''' 2 and 9 ... '''là''' and '''tú'''. When fitted to '''maŋga''' they produce an adjective and an adverb respectively.


'''tore doikatu''' = "he/she came on foot" or  "he/she came by walking"
..
 
We'll talk about and adverbial construction first ...


'''tore doikatu''' = "he/she came on foot" or  "he/she came by walking"
'''tore doikatu''' = "he/she came on foot" or  "he/she came by walking"
Line 943: Line 965:
'''tore tú doika saco''' = "he/she came by walking quickly"
'''tore tú doika saco''' = "he/she came by walking quickly"


This "'''tú'''-construction" acts as an adverb. Notice that the particle '''tú''' acts as it normally does and appends to the end of a single word, but stands alone to the left of a multi-word phrase.
Notice that the particle '''tú''' acts as it normally does and appends to the end of a single word, but stands alone to the left of a multi-word phrase.
 
..
 
And the adjectival constuction ...


..
..
..
{|
|-
! bàu || doika-la
|
| man || walk-{{small|1SG-IND-PRES}}
|}
{|
|-
! bàu || r  ||doika-la
|
| man || {{small|COPULA} ||walk-{{small|P2}}
|}


Also '''là''' appears often in conjunction with '''manga'''
Also '''là''' appears often in conjunction with '''manga'''
Line 977: Line 1,019:
..
..


Also '''''' appears often in conjunction with '''manga'''
'''maŋga''' ... as well as appearing as arguments in a clause. That is S, A, O, CS and CO, also appear as complements to auxiliary verbs.


The ''''''-constuction acts as an adverb. An adverb meaning "the '''r'''-form (matrix verb) happened during the time of the action qualified by ''''''.
One such auxilliary is '''tuma''' meaning "to squeaze" or "to force". [ when it means "to squeaze" it is followed by a '''kenʒi''' and is acting as a normal verb, when it means "to force" it is followed by a '''manga''' and is acting as an auxiliaryl verb ]


'''jonos lailore doikapi''' = "John sang while walking earlier today"


'''jonos lailore pí doika tunheun''' = "John sang while walking to the civic centre earlier today"
In these constructions, there is a very strong tendency (almost a rule) that the '''maŋga''' is on RHS of  '''maŋguma'''. However if an ergative (A) argument is present, that element can instead be on RHS. Also a very strong tendency fot the O aurgment to directly follow the '''maŋga'''. There is a fairly strong tendency for the S argument to directly follow the '''maŋga'''.


..
So ...


'''manga''' ... as well as appearing as arguments in a clause. That is S, A, O, CS and CO, also appear as complements to auxiliary verbs.
Thomas forced John to hit Jane => '''tomos tumori timpa jene hí jono''' or '''tomos hí jono timpa jene'''


One such auxilliary is '''tuma''' meaning "to squeaze" or "to force". [ when it means "to squeaze" it is followed by a '''senko''' and is acting as a normal verb, when it means "to force" it is followed by a '''manga''' and is acting as an auxiliaryl verb ]
Thomas forced Jane to walk => '''tomos tumori doika jene'''


'''tomos tumori doika jene''' = Thomas forced Jane to walk .... [ note '''doika jene''' is one element and must stay in this order ]
[By the way ... as an example of '''tuma''' being a normal verb ... '''tomos jwuba komo jene tumori''' = Thomas squeazed Jane's left buttock ]


'''tomos tumori timpa jene hí jono''' = Thomas forced John to hit Jane ... [ note '''timpa jene hí jono''' is one element an must stay in this order ]
..
 
[Note to self : think about the below more ]. These two expressions also have alternatives ...
 
'''tomos tumori jene doika'''


'''tomos tumori jono timpa jene'''  
Two other examples of '''maŋga''' with auxilliary verbs (why not) ...


[By the way ... as an example of '''tuma''' being a normal verb ... '''tomos jwuba komo jene tumori''' = Thomas squeazed Jane's left buttock ]
1) ... '''mbe''' = to hold ..... '''lelpa''' = to sing, singing ..... '''jenes mbor lelpa bòi''' = Jane can sing well. ['''lelpai''' = a song ?]
 
Two other examples of '''manga''' with auxilliary verbs (why not) ...
 
1) ... '''mbe''' = to hold ..... '''lelpa''' = to sing, singing ..... '''jenes mbor lelpa bòi''' = Jane can sing well.


2) ... '''glù''' = to depart ... '''timpa''' = to hit, hitting ...  '''jonos glori timpa jene''' = John stopped hitting Jane
2) ... '''glù''' = to depart ... '''timpa''' = to hit, hitting ...  '''jonos glori timpa jene''' = John stopped hitting Jane
Line 1,011: Line 1,044:
..
..


One notable use of the '''manga''' is emphasis, where the '''manga''' is used right next to the same word in '''r'''-form. For example ...
One notable use of the '''maŋga''' is emphasis, where the '''manga''' is used right next to the same word in '''r'''-form. For example ...


{|  
{|  
Line 1,019: Line 1,052:
| die-{{small|3SG-IND-FUT}} || death
| die-{{small|3SG-IND-FUT}} || death
|} = He/she will die a death => He/she will die for sure
|} = He/she will die a death => He/she will die for sure
..


{|  
{|  
Line 1,026: Line 1,061:
| live-{{small|3SG-IND-PAST}} || life
| live-{{small|3SG-IND-PAST}} || life
|} = He/she lived a life => He/she had a full life
|} = He/she lived a life => He/she had a full life
..


{|  
{|  
Line 1,035: Line 1,072:




Now '''maumori mauma''' and '''daw.oru dàu''' are strange. Both verbs are strictly intransitive. If the pronouns were to be included you would have to absolutive arguments in the clause (i.e. '''ò''' and '''dàu''' : '''pà''' and '''mauma''') ... something that is not allowed. But it seems that '''béu''' does allow it if one of the arguments is the base of the '''r'''-form.
Now '''maumori mauma''' and '''daw.oru dàu''' are strange. Normally both verbs are strictly intransitive. But here there are transitive. Seemingly sometimes '''béu''' allows
 
a) intransitive => transitive
b) A argument => absolutive form
c) O argument = '''maŋga'''  
 
'''pà maumari mauma''' is the answer to '''mìn maumari mauma'''  


..
..
..


----
'''*''' Another way to say this is '''solbe saco moze**'''
Note to self : Dixon makes a big deal over the below. [ Note to self : what sort of deal should I make ? ]
 
1) The killing of the president was an atrocious crime. [ Note to self : "killing" here is a noun ... as identified by "the" and "of the" ]


2) Killing the president was an atrocious crime. [ Note to self : "killing" here is not a noun ... we can call it an infinitive ... as identified by "the president" following immediately ]
'''**''' If '''saco''' doesn't immediately follows the '''manga''', it must be explicitly tagged as an adverb by the -'''is''' suffix. The exact same rule as for (finite)clauses.


You can see that one form "killing" is used in 2 different constructions.  
----
Note ... "I saw a man cutting the grass" is an English clause. I think Dixon analyses "the man cutting the glass" as a complement clause ??? This sees a bit strange to me. The '''béu''' equivalent .... '''hecari bàu katala lazde''' is just analyzed as Verb '''hecari''' ... Object '''bàu''' and Adjective Phrase '''katala lazde'''


And a further note ... "I saw a man cutting the grass" is an English clause. I think Dixon analyses "the man cutting the glass" as a complement clause ??? This sees a bit strange to me. The '''béu''' equivalent .... '''mwari bàu katala lazde''' is just analyzed as Verb '''mwari''' ... Object '''bàu''' and Adjective Phrase '''katala lazde'''
----
----


Line 1,063: Line 1,104:
Here, for a transitive clause, "that which initiates the action" is called the A argument, and "that which is affected by the action" the O argument. Also, for an intransitive verb, the noun is called the S argument. It is convenient to make a distinction between all three cases. I follow RMW Dixon in using this terminology.  
Here, for a transitive clause, "that which initiates the action" is called the A argument, and "that which is affected by the action" the O argument. Also, for an intransitive verb, the noun is called the S argument. It is convenient to make a distinction between all three cases. I follow RMW Dixon in using this terminology.  


In most languages the S argument is marked the same way as the A argument. However in some languages the S argument is marked the same way as the O argument. These are called ergative languages. '''béu''' is one of these ergative languages. About a quarter of the world languages are ergative or partly ergative.  
In most languages the S argument is marked the same way as the A argument. However in some languages the S argument is marked the same way as the O argument. These are called ergative languages. '''béu''' is one of these ergative languages. About a quarter of the world languages are ergative or partly ergative. The ergative system ...
 
Below are the '''béu''' pronouns for the S and O arguments. This form can be considered the "unmarked form".  


..
..


{| border=1
{|  
  |align=center| me
|-
  |align=center| '''pà'''
! bàu || || glá || maum-u-r-i
  |align=center| us
|-
  |align=center| '''magi'''
| man || and || woman||align=center| sleep-{{small|3PL-IND-PAST}}
  |align=center| inclusive
|} ==> The man and the woman slept
  |-
  |align=center|
  |align=center|  
  |align=center| us
  |align=center| '''manu'''
  |align=center| exclusive
  |-
  |align=center| you
  |align=center| '''gì'''
  |align=center| you
  |align=center| '''jè'''
  |-
  |align=center| him, her
  |align=center| '''ò'''
  |align=center| them
  |align=center| '''nù'''
  |-
  |align=center| it
  |align=center| '''ʃì'''
  |align=center| them
  |align=center| '''ʃì'''
    |}


..
NOTE ... Pronouns  differ from nouns in that their tones change between the ergative and the unmarked form. For a normal noun it is sufficient that -'''s''' is suffixed. For example ...
..


{|  
{|  
Line 1,108: Line 1,120:
! bàu-s ||align=center| glá || timp-o-r-i
! bàu-s ||align=center| glá || timp-o-r-i
|-
|-
| man-{{small|ERG}} |align=center|woman||align=center| hit-{{small|3SG-IND-PAST}}
| man-{{small|ERG}} || align=centre|woman|| hit-{{small|3SG-IND-PAST}}
|} ==> The man hit the woman
|} ==> The man hit the woman


Line 1,116: Line 1,128:
! bàu ||align=center| glá-s || timp-o-r-i
! bàu ||align=center| glá-s || timp-o-r-i
|-
|-
| man ||align=center| woman-{{small|ERG}}|| hit-{{small|3SG-IND-PAST}}
| man ||align=centre| woman-{{small|ERG}}|| hit-{{small|3SG-IND-PAST}}
|} ==> The woman hit the man
|} ==> The woman hit the man


..
..


Below are the pronouns in the ergative form.
Below are the '''béu''' pronouns for the S and O arguments. This form can be considered the "unmarked form".  
From now on I will call the ergative form the '''s'''-form, and the unmaked form the base form.


..
..


{| border=1
[[Image:SW_119.png]][[Image:SW_108.png]]
  |align=center| I
  |align=center| '''pás'''
  |align=center| we
  |align=center| '''magis'''
  |-
  |align=center|
  |align=center|
  |align=center| we
  |align=center| '''manus'''
  |-
  |align=center| you
  |align=center| '''gís'''
  |align=center| you
  |align=center| '''jés'''
  |-
  |align=center| he, she
  |align=center| '''ós'''
  |align=center| they
  |align=center| '''nús'''
  |-
  |align=center| it
  |align=center| '''ʃís'''
  |align=center| they
  |align=center| '''ʃís'''
    |}


..
..


'''''' and '''jés''' are the second person plural forms.  
'''mpau''' includes the listener in the "we", '''pài''' excludes the listener from the "we" and '''''' indicates a plural "you".


There is one other pronoun ... the reflexive pronoun ''''''. This is always an O argument. Notice that it is the only O argument with a high tone.
The proper way to use '''?o''' is to use it for all non-humans. But in actually practice, some people might "elevate" the status of a family pet and call it '''òn''' or '''ʃì'''.


'''*''' In the Western Linguistic Tradition, these "forms" are called "cases". The English word case used in this sense comes from the Latin casus, which is derived from the verb cadere, "to fall", from the Proto-Indo-European root *ḱad-.  The Latin word is a calque of the Greek πτῶσις, ptosis, "falling, fall". The sense is that all other cases are considered to have "fallen" away from the nominative (considered the unmarked form in Latin).
'''onde''' is used for "they" when they are all male, '''ʃide''' is used for "they" when they are all female, '''ʃinde''' is used for "they" when they are mixed, and '''?ode''' is used for "they"when they are all non-human.
 
'''**''' By the way, there are 17 marked forms (cases) in '''béu''' ... the ergative being one of these.
 
..
 
== ..... Intensifiers==


..
..


THIS MUST BE REWRITTEN .... TUGE = more : JIGE = less
Below are the '''béu''' pronouns for the A argument  ... the "ergative form".
 
Remember earlier in this chapter, we mentioned the numerative slot (for the '''senko'''). To recap, this slot can contain ...


'''yè''' "plural" ... '''aʔa''' "one" ... '''ima''' "two" ... '''uya''' "three" ... '''iyo''' "few" ... '''eja''' "four" ... '''ofa''' "five" ..... up to ..... '''afaufaifa''' "215<sub>10</sub> ... '''hài''' "many"and '''ú''' "all"


Below is show how '''hài''' and '''iyo''' divide up the semantic space of quantity(intensity).
[[Image:SW_120.png]]


..
..


[[Image:TW_788.png]]
NOTE ... Pronouns differ from nouns in that their tones change between the ergative and the unmarked form. For a normal noun it is sufficient that -'''s''' is suffixed. For example ...


..
From now on I will call the ergative form the '''s'''-form, and the unmaked form the base form.
 
Now all '''saidau'''(adjectives) can be affixed by -'''ge''' to form the comparative'''*''' form. For example ...
 
'''bàu jutu''' = "the big man" : '''bàu jutuge''' = "the bigger man"
 
This affix can also be used with the numbers ...
 
'''juge''' "more than zero", '''a?age''' "more than one" : '''image''' "more than two" .... up to '''afaufaifage''' "more than 215<sub>10</sub>'''**'''
 
Now -'''ge''' can also be affixed to '''iyo''' letting us fill in every box of the chart given above ... [[Image:TW_789.png]]


..
..


Now when attached to '''saidau''', -'''ge''' gives a relative value (i.e. you are comparing one thing with another). However when -'''ge''' is attached to a numbers you get an absolute value (i.e. you are not comparing the modified item with anything).
There is one other pronoun ... the reflexive pronoun ''''''. This is always an O argument. Notice that it is the only O argument with a high tone.
 
When you want to compare two items as to their numerative value, you must use the particle '''yú'''.
 
(The word '''yú''' and the suffix -'''ge''' both can be translated as "more", however '''yú''' only qualifies nouns and -'''ge''' only qualifies adjectives)
 
'''jonos byór yú klogau jenewo''' = "John has more pairs of shoes than Jane"
 
'''?ár yú halmai''' = "I want more apples"
 
'''?ár hài halmai''' = "I want a lot of apples" or "I want many apples"


..
..


Now a number can immediately follow ''''''. For example ...
'''*''' In the Western Linguistic Tradition, these "forms" are called "cases". The English word case used in this sense comes from the Latin casus, which is derived from the verb cadere, "to fall", from the Proto-Indo-European root *ḱad-. The Latin word is a calque of the Greek πτῶσις, ptosis, "falling, fall". The sense is that all other cases are considered to have "fallen" away from the nominative (considered the unmarked form in Latin).  


'''?ár yú léu halma''' = "I want three more apples"
'''**''' By the way, there are 17 marked forms (cases) in '''béu''' ... the ergative being one of these.
 
'''yár yú halmai jenewo''' = "I have more apples than Jane" ....... [ note ... '''halma''' with '''léu''' but '''halmai''' with '''yú''' ]
 
..
 
To indicate "less" ... use '''wì'''. For example ...
 
'''jenes yór wì halmai pawo''' = "Jane has less apples than me"
 
'''jenes yór wì hói halma pawo''' = "Jane has two less apples than me" .... but it would sound better to rephrase these as ...
 
'''yár yú halmai jenewo''' = "I have more apples than Jane" : '''yár yú hói halmai jenewo''' = "I have two more apples than Jane"
 
..
 
'''*'''The affix -'''mo''' is the superlative for adjectives. When joined to '''hài''' and '''iyo''' ... we get "the majority" '''haimo''' and "the minority" '''iyomo'''
 
'''**'''Note ... the words '''noge''', '''haige''' and '''uge''' do not exist.
 
..
 
----
 
..
 
Above we have talked about numeratives and in detail about how to quantify '''senko'''.
 
Below we will touch on how other categories of words have their own intensifiers ...
 
..
 
[[Image:TW_920.png]]
 
..
 
'''hài bàu''' = many men
 
'''moze hè''' = a lot of water
 
 
'''hè''' also can qualify verbs. As with normal adverbs, if it doesn't immediately follow the verb it must take the form '''hewe'''.
 
(Note to self : I can't think of a reason you would want to separate '''hè''' from its verb)
 
'''glá doikori hè''' = the woman walked a lot
 
'''hewe glá doikori''' = the woman walked a lot
 
'''báus timpori glá hewe''' = the man hit a woman a lot
 
And also can intensify '''manga''' and '''mangas'''
 
'''solbe hè moze''' = "to drink a lot of water"
 
'''solbe moze hè''' = "to drink a lot of water"
 
The above two forms are equally likely to be found. There is a difference in meaning but you would be a real nitpicker to worry about that.
 
..
 
'''saidau''' and '''saidaun''' are both intensified by '''sowe''' ...
 
'''jutu sowe''' = "very big"
 
'''jutun sowe''' = "the very big one"
 
..
 
Notice that '''mangan''' and '''saidaun''' can take two intensifiers ...
 
'''hài solben hè wiski''' = the many times a lot of whisky was drink ... '''hài solben hè wiski hí pà''' = the many times I have drunk a lot of whisky
 
'''hài gèun sowe''' = the many very green ones
 
..
 
We will take about the opposite of intensifiers and other quantifiers in a later chapter. These are a lot rarer. The intensifiers are the ones most commonly used.
 
..
 
== ..... The main derivation pathways==
 
..
 
Derivational morphology often involves the addition of a derivational suffix or other affix. Such an affix usually applies to words of one lexical category (part of speech) and changes them into words of another such category. For example, the English derivational suffix -ly changes adjectives into adverbs (slow → slowly).
 
Examples of English derivational patterns and their suffixes:
 
*adjective-to-noun: -ness (slow → slowness)
*adjective-to-verb: -ize (modern → modernize)
*adjective-to-adjective: -ish (red → reddish)
*adjective-to-adverb: -ly (personal → personally)
*noun-to-adjective: -al (recreation → recreational)
*noun-to-verb: -fy (glory → glorify)
*verb-to-adjective: -able (drink → drinkable)
*verb-to-noun (abstract): -ance (deliver → deliverance)
*verb-to-noun (agent): -er (write → writer)
 
Derivation can be contrasted with inflection, in that derivation produces a new word (a distinct lexeme), whereas inflection produces grammatical variants of the same word.
 
Generally speaking, inflection applies in more or less regular patterns to all members of a part of speech (for example, nearly every English verb adds -s for the third person singular present tense), while derivation follows less consistent patterns (for example, the nominalizing suffix -ity can be used with the adjectives modern and dense, but not with open or strong).
 
Derivation can also occur without any change of form, for example telephone (noun) and to telephone. This is known as zero derivation. [ All the above from "wikipedia" under "linguistic derivation" ]
 
..
 
The diagram below shows the ten main derivational processes which are absolutely fundamental to the working of the language.
By the way, the verbal base (equivalent to English infinitive or gerundive I think ??) should be considered a noun.
 
 
[[Image:TW_917.png]]
 
 
[1]
 
Most nouns can be used as adjectives just by placing them directly after the noun they are qualifying. Like "school bus" in English. For example ...
 
'''pintu tìa''' = a/the door of the house
 
Also to indicate possession the possessee is usually just placed after the possessed.
 
'''tìa jono''' = John's house
 
(Actually there is a particle '''yó''' joining the possessed to the possessee ... however it is rarely used. '''yó''' is also a noun meaning possessions, '''yái''' an item possessed, '''yáu''' "to have")
 
"John's house" => '''tìa yó jono''' .... but more usually '''tìa jono'''
 
This is zero derivation and is marked as [[Image:TW_816.png]] in the above diagram.
 
[2]
 
'''gèu''' = green
 
'''+ gèu''' = the green one
 
'''?azwodus''' = lactose intolerant
 
'''+ ?azwodus''' = a/the lactose intolerant one
 
[3]
 
'''gèu''' = green
 
'''k+ gèu''' = the green ones
 
'''k+ gèu làu oila''' = six green ones
 
'''sadu''' = elephant
 
'''k+ sadu''' = elephant-kind
 
'''k+ sadu làu oila''' = six elephants ... well, it is legitimate to say this ... but '''oila sadu''' is so easier.
 
[4]
 
'''gèu''' = green
 
'''kuwai gèu''' = greenness
 
[5]
 
'''yubau''' = strong
 
'''yubako''' = to strengthen
 
'''pona''' = hot
 
'''ponako''' = to heat up
 
[6]
 
'''poma''' = kick (also means leg) .... '''pomora''' = He/she is kicking
 
'''pomako''' = to kick
 
However if the base noun ends in '''n''' ...
 
'''kwofan''' = bicycle
 
'''gàu kwofan''' = to (do) bicycle
 
[7]
 
'''pazba yubara'''  "I am strengthening the table"
 
..
 
{|
|-
! pazba || yub-a-r-a
|-
| table || strengthen-{{small|1SG-IND-PRES}} 
|}
 
'''ponara moze'''  "I am heating up some water"
 
{|
|-
! pon-a-r-a || moze
|-
|  "heat up"-{{small|1SG-IND-PRES}} || water
|}
 
[8]
 
'''tunheun kwofanaru''' "I will bicycle to the townhall"
 
..
 
{|
|-
! tunheu-n || kwofan-a-r-u
|-
| townhall-{{small|DAT}} || bicycle-{{small|1SG-IND-FUT}} 
|}
 
[9]
 
This will be covered in detail in the next chapter. However here is a quick example ...
 
'''solbara moze'''   "I am drinking water"
 
..
 
{|
|-
! solb-a-r-a || moze
|-
| drink-{{small|1SG-IND-PRES}} || water
|}
 
from the verb base '''solbe''' "to drink"
 
[10]
 
-'''s''', -'''n''', -'''a''', -'''o''' take -'''is''', all other endings take -'''s''' (including -'''ia''' and -'''ua''')
 
'''saco''' = fast, '''sacois''' = quickly
 
'''pudus''' = timid (of an animal), '''puduʒis''' = timidly
 
'''yubau''' = strong, '''yubaus''' = strongly
 
..
 
.
 
For [7] and [8] if the root that is to be transformed is monosyllabic, then we need -'''ko''' as well as -'''r'''-. For example ...
 
..
 
'''bàu''' = man
 
'''bauko''' = to man (exact same meaning as in English)
 
'''baukara téu dí''' = I am manning this position.
 
..
 
'''gèu''' = green
 
'''geuko''' = to make green
 
'''geukara pazba dí''' = I am painting this table green
 
..
 
You can say, that for monosyllabic words [7] = [5] + [9] and  [8] = [6] + [9].
 
 
..
..
 
 
Unadorned adjective can be used as nouns in many situations. Similar happens in many languages. For example ... '''klár gèu''' is ambiguous.
 
To disambiguate => '''klár kuwai gèu''' "I like greenness"  /  '''klár k+ gèu''' "I like the green ones" /  '''klár + gèu''' "I like the green one"
 
 
 
 
 
 
.
----
 
The remaining two transformations shown on the diagram are for verbalization. Actually the affix -'''ko''' is added to all adjectives or nouns in order to make a verb. However in one circumstance this affix is not needed. This is for the '''r'''-form based on a multi-syllable adjective or noun. For example ...
 
..
 
'''pazba yubaku''' = strengthen the table (a command)
 
'''pazba yubakis''' = you should strengthen the table
 
..
 
'''ponaku moze''' = heat up some water  (a command)
 
'''ponakos moze''' = he/she should heat up some water
 
..
 
 
 
 
'''bauku téu dí''' = man this position  (a command)
 
'''baukos téu dí''' = he/she should man this position
 
 
 
 
 
'''naike''' = sharp : '''naikeko''' = to sharpen
 
'''keŋkia''' = salty : '''keŋkiko''' = to add salt  ... when the adjective ends is a diphthong (and is non-monosylabic) the last vowel is dropped.
 
'''keŋkikara''' = "I am adding salt"  .... note <u>not</u> '''*keŋkara''' ... this is because '''keŋkia''' is a derived word.
 
'''sài''' = colour : '''saiya''' = colourful : '''saiwa''' = colourless : '''saiko''' = to paint (maybe via '''*saiyako''')
 
..
 
Note ... -'''ko''' is possibly an eroded version of '''gàu''' ( "to do" or "to make" ).
 
Note ... There seems to be a method of deriving a two place verb from a one place verb by affixing -'''n'''. For example ... '''diadia''' = "to happen" : '''diadian''' = "to cause". While this mechanism is seen all over the language I have not mentioned it in the chart above. This is because I consider it non-productive. I count '''daidia''' and '''diadian''' both as base words. In a similar way that English speakers consider "rise" and "raise" independent words, "lie" and "lay" independent words and "sit" and "set" independent words.


..
..

Latest revision as of 21:00, 26 June 2020

TW 415.png

..... The 5 basic word types

..

All words are derived from these 5 basic types. Actually the fengi "particle" have so many subtypes (often single particles are a subtype to themselves) that it is a bit of a fudge to say that béu has 5 basic types. Maybe more honest to say that béu has 4 groups of words and the behaviour (syntactically) of any word in these 4 groups depends on which group it is in.

..

1) fengi = particle ... this is a sort of "hold-all" category for all words (and affixes) that don't neatly fit into the other categories. Interjections, numbers, pronouns, conjunctions, determiners and certain words that would be classed as adverbs in English, are all classed as fengi.

By the way ... all affixes are counted as a type of fengi.

An example is .. the preposition indicating the oblique case.

..

2) kenʒi = an object

An example is bàu ... "a man"

..

3) olus = material, stuff

An example is moze ... "water"

..

4) saidau = adjective

An example is nelau ... "dark blue"

..

5) manga = a verb in its base form (citation form). When used "actively" it will take its r-form, u-form or i-form.

An example is twá meaning "to meet" or "a meeting" (the concept of "meet" disassociated from any arguments, tense, aspect or whatever).

..

SW 062.png

..

..... Kenʒi

..

kenʒi can mean "noun". It can also mean "noun phrase" (NP).

.

Probably the most "basic" of the basic 5 ... tangible and discrete.

The noun can take six types of modifiers. These six types must come in a certain order ...

..

SW 070.png

..

In the above diagram, an descending arrow followed by a bar indicates a closed set. A descending arrow by itself indicates an open set. Branching arrows indicate multiple possibilities.

The head of the NP can be referred to as kenʒita. Usually it is called this by lay people and by linguists when the concept is first brought up. However, thereafter it is usually referred to as húa meaning "head".

kenʒita is kenʒi plus the diminutive suffix. kenʒi can also take the augmentative suffix -uma. kenzuma "extended noun phrase" is a normal kenʒi, with either a relative clause (RC) appended to the right or a partitive appended to the left hand side

The words highlighted in red convert the noun phrase (or indeed the sentence in which the NP is embedded) into a question. A blue circle indicates the only mandatory element. But even these elements can be dropped on occasion ... when they are understood from context or the preceding conversation. When we have one adjective, and the head is understood, ɘ can be substituted for the head, if the head is plural.

ɘ gèu = a/the green one : kɘ gèu = a/the green ones

These two particles can also be used with other noun modifiers, however not always mandators with non-adjective modifiers.

ɘ nái = which one : kɘ nái = which ones

kɘ dí = these ones : ɘ dè = this one

However nái, and can constitute NP's by themselves. A bit like English

Looking at the chart above might give you a false impression of béu noun phrases. The number of modifiers within a noun phrase is usually only one or two. When there is two, they must occur in a certain order, hence the necessity of the chart above. I don't think it would be easy to process a noun phrase with six modifiers, probably some of them would be shunted off into a RC with an initial copula. A noun phrase can take multiple RC's. They can stand beside each other in a sort of apposition.

I should make one further point here. The particles ú "all" and "no" can appear to the left of the head. They can also appear in the quantity slot.

..

... Quality

..

More than one adjective is allowed in this slot. For example ... bàu gèu tiji = the little green man

kái meaning "what type" can also appear in this slot. In which case it turns the whole noun phrase (or sentence) into a question. For example ...

bàu gèu kái = what kind of green man ? ... (NP question)

há bàu gèu kái glà timpori = what kind of green man hit the woman ? ... (sentence question)

Now when you have multiple adjectives they will have a certain order depending upon their sub-category.

This is the same as English ... for example, you always say "the third big black dog" and never "the black third big dog".

béu uses the exact same order as in English but the other way around.

béu has two adjectives that come in this slot that are worth mentioning. They might have claims to particle-hood, but I guess their appearance in this slot marks them as adjectives. No reason that they can't be both.

1) ... = "other"

The semantics of this word remind me of the semantics of tuge/jige. With the relative quantifiers the speech participants have agreed on the number/amount relevant to the situation. tuge/jige are used to change this value. Similarly is used in a situation where the speech participants have agreed on the population (of whatever noun category) under consideration but one of them wants to expand this population.

2) ... laubo = enough

..

... Quantity

..

This slot is very interesting ...

SW 071.png

The above chart is split into definite and vague sections. All the items under definite represent an integer (or "the empty set" or "the full set"). The items under vague represent an approximate number/amount. This section is further divided into discrete and non-discrete (i.e. countable.non-countable).

modifies both discrete and non-discrete. It means a moderate amount ... some value between zero and "all". It does NOT mean "indefinite" ... "some man" is bàu èn, not *bàu yè.

This word can be used to mark plurality (together with iyo and hài) for those nouns that can not be pluralized in themselves. For example ... húa, "head" : húa yè, "heads".

jí jí and jía are about equally common and mean the same thing. However jía tends to be used in more formal situations and jí jí in less formal.

..

láu (how many) can appear in this slot. In which case it turns the NP (and hence the clause containing the NP) into a question. For example ...

bàu jutu láu = "How many men are big ?" or "How many big men ?" ???????????

..

The chart above shows only the terms used for absolute quantity ????????????????? It does not cover, what I call relative quantity. Let me explain ...

Imagine the speaker and the hearer both have an idea of the number/amount relevant to a situation but one of them wants to change this number/amount. The amount he wants to change this agreed number/amount by, I call the relative quantity. It can be positive or negative. When positive we use the word tuge "more" ... when negative we use the word jige "less" *. For example ...

turi waudo tuge = more dogs came

t-u-r-i waudo tuge
come-3PL-IND-PST dog more

These to particles can be modified by some (most) of the terms given in the chart above. They can be modified by any of the terms hi-lighted in orange.

Sw 060.png

For example ... bía tuge ima = two more beers please"

Note : actually jía and tundu are not applicable to kenʒi. They are only applicable to olus

..

* These words might be derived somehow from jutu "big" and tiji "small" ... along with the comparative suffix -ge **.

The comparative suffix can be appended to any adjectives. For example ... jini "clever" => jinige "cleverer" : hau?e "beautiful" => hau?ege "more beautiful"

There is also a superlative suffix ... -mo. So jinimo "cleverest" amd hau?emo "most beautiful"

** There is an independant word which might be related to the comparative suffix. It is a particle that always comes in twos. For example ... gé tundu ... gé bói "the more the merrier".

Sometimes you coma across bù tuge "no more". This should be analysed as a contraction of bù ?ár tuge "I don't want more".

*** Perhaps wóin is related to the verb gwói "to pass by" plus the past participle -in.

..

... Ownership

..

Basically you can just stick a personal name, a pronoun or any NP in here and the head noun will be considered owned by the object inserted here.

Sometimes, the particle precedes the object inserted.

For example jwado gèu yó jene = Jane's big green bird

Note that the particle is usually dropped when the possessor is next to the head. However as other elements intervene, the likelihood that is used increases.

If mín (who) is stuck in this slot ... then we have a question. For example ...

jwado gèu yó mín = Whose big green bird ? = Whose's the big green bird ?

There can be ambiguity with some kenʒi possessing a genitive. For example ...

Does waudo bàu dí mean "the dog of this man" or "this dog of the man" ?

To get around this, we have a special rule ...

"If anything is in the ownership slot, and never appear in the determiner slot. Instead they appear as dían "here" and dene "there" in the locative slot"

Note ... sometimes ownership as such is not what is of interest, it is if a person has actual physical possession. In this case is not used. But the object takes pila?o 2.

jwado gèu là Long John Silver catora = The big green bird (on Long John's shoulder presumably) is chatting away.

Actually segments showing actually physical possession like the example above, go in the locative slot which we will cover next.

..

... Location

..

Ordinal numbers appear in this slot. The ordinal numbers are ...

SW 066.png

You will notice that there are two words for first ... da?a and dahua. They are both equally common, but da?a tends to occur in the presence of dima or duya while dahua tends to occur in the presence of dauci.

..

Proper locatives comprise a noun plus one of the 9 pila?oi .... pi la mau goi ce do bene komo ni. For example ...

duzu pobomau = The oryx on the mountain

Also pila?o 14 turns up in this slot. These items are strictly not giving information about "location" but rather "origin". They are classed as a locatives nevertheless. For example ...

bàu glazgofi = a/the man from Glasgow

If the location consists of more than one word, the usual rule applies and the pila?o appears as a preposition ...

duzu máu pobo jutu = The oryx on the big mountain

There is a tendance that the longer the locative item, the more likely the locative item will be shunted into a relative clause ...

duzu nài r máu pobo hau?e jutu = The oryx on the big beautiful mountain

nài r máu pobo hau?e jutu is a relative clause. We will cover RC's in a bit.

All prepositions that are not pila?o lead to the location being shunted into a relative clause. For example ...

polga?o nài r fiagan gwai = "the sailing boat which is among the islands" or simply "the sailing boat among the islands"

..

Also "where" can appear in this slot. In which case it turns the noun phrase into a question. For example ...

bàu gèu dá = where is the green man ?

..

Addendum ... To keep things simple I refrained from mentioning this before but time can also be expressed in this slot. For example "the turmoil 1300's". Actually location AND time can both be in this slot at the same time. For example "the anger Iceland 1950's ...".

..

* Probably derived from uci "tail".

..

... Determiner

..

There are five of these ... (this), (that), nái (which), èn (some) and ín (any) . For example ...

and are called demonstratives in the WLT. They will be covered in the section after next.

nái turns the whole noun phrase into a question. For example ...

bàu gèu tiji nái = which little green man ? ... noun phrase question

And of course, if a NP represents a question, any clause containing this NP will also be a question. For example ...

bàu gèu tiji nái glà timpori = which little green man hit the woman ? ... a clause AND a question

èn "some" appear in this slot ... bàu gèu tiji èn = "some little green man" ...... indefinite

ín "any" appear in this slot ... bàu gèu tiji ín = "any little green man" .............. super indefinite

There is one little rule to remember ...

"Only one item is allowed in this slot, so if you want an indefinite as well as a demonstrative, the demonstrative is shunted off to the locative slot and given the form dían or dene."

I guess this is logical in a way. and were originally associated with pointing. But when the object is indefinite, how can you point ? "here" or "there" is about as definite as you can get.

..

... Side-note re demonstratives

..

"this" and "that" are two words that orientate and focus the hearer's attention on an object (or location *) in the speech situation. These words are called demonstratives in the WLT.

According to Holger [ Diessel (1999:57) ] ...

i) A demonstrative can be construed as an argument in its own right. That is, it can constitute a NP without any additional elements.

ii) A demonstrative can co-occur with a noun in a NP. That is, it can be a noun modifier.

iii)* A demonstatives can function as a verb modifier. It specifies (the) location (where something happens **).

* Perhaps in a more earlier version of the WLT "location" and (iii) would not be included in the definition of determiner. English and béu conform to this earlier version of the WLT. However I think it is a good idea when considering all the world's languages, to use this wider definition of "demonstrative".

"**" Perhaps in a language where a copula is not routinely used "where something happens" would not necessarily be appropriate.

And here are examples of the above three functions (in English) ...

a) This is excellent.

b) That guy is an idiot.

c) Here we do things differently.

Diachronically, these three functions can run into each other. Function (a) and function (b) are particularly close. They have the exact same form in English, but no confusion can occur, because "this/that:b" can be deduced to be inside a NP by the rules of English grammar. Most languages in the world (70%) have identical forms for "this/that:a" and "this/that:b". Of the languages that do not have identical forms, the difference can be quite subtle. For example in Thai นี่ [ nii falling tone ] is "this:a" and นี้ [ nii high tone ] is "this:b". ........... [see WALS 42A]

Some languages lack (a). For example, in Korean, to express "this:a" you must say "ce il" meaning "this thing". So (b) used instead of (a)

Some languages lack (b). They would say something like "the guy here" instead of "this guy". So (c) instead of (b)

Some languages lack (c). They would say something like "this place we do things differently" instead of "here we do things differently". So (b) instead of (c.)

[ And while we are talking on this area, perhaps we should mention 3rd person pronouns (see WALS 43A). Some languages lack 3rd person pronouns. They cover this function by saying something like "this" or "that guy" ... A further point of interest (well, I find it interesting anyway) is that the English he and here are cognates. Going back to a P.I.E. form meaning (a) or (b). -r was a ProtoGermanic adverbial suffix. ]

béu patterns pretty much like English (and the pattern of English is not atypical of the world's languages) ...

= "this:a" : = "that:a"

= "this:b" : = "that:b"

dían = "this:c" (i.e. "here") : dene = "that.c" (i.e. "there")

I was originally thinking of just appending the béu adverbial suffix -is to produce (c). But rejected that idea in order to get more phonological contrast between ...

(A) "this:c" and "that.c", (B) "this/that:a/b" and "this/that:c"

With dían there is a hint that it might be derived from plus pila?o 15. And also with dene ... a hint that it might have the same origin. But who can tell. These things are lost in the mists of time.

..

... Further uses of and

..

If we first hear a plural noun articulated in a conversation, the most likely meaning we would assigned to it would be the universal set. For example moltai.a. There is a more explicit means to express the universal set. For example ... kài moltai = "doctor.kind" but this construction is seldom used.

An example of usage is ... moltai.a súr jini = "doctors are clever"

OK ... now lets zoom in a bit. To zoom in we need to take in or give out some narrative. So now we hear the following ....

Next week British junior doctors will withhold many services in protest against the long hour expected of them

OK ... after hearing that ... moltai.a dè would be taken to mean "British junior doctors"

OK ... lets hear a further bit of narrative ...

Much to the disgruntlement of the senior doctors who will have a hard week ahead of them making up for the short fall.

OK ... after hearing that ... moltai.a dè would be taken to mean "British senior doctors". So, what refers to doesn't persist long, Our perspective is continually changing.

[ I can't help thinking that the proximate/obviate system existing in Plains Cree would be very useful. You could keep track of two protagonists through a discourse without reverting to full NPs. But I guess there are cognative reasons why it is difficult to use. Well, if it was easy to use, it would be far more wide-spread. It must be very useful. ]

This is in normal discourse. However if some objects are physically pointed out * when first introduced (and presumably they stay in sight for the duration of the discourse) what and referred to would persist.

So we can see that points back in time. It brings to the top of consciousness, the last set of doctors talked about.

..

In a narrative many objects are encountered. If a newly introduced object is marked by it means that the object is important to the narrative and you will shortly be getting more information about it. The process is not exactly the inverse of anaphora. But one is compatible with "information given in the past leading to easy identification of which object in particular we are talking about. The other is compatible with "in the near future I will give you information about this object and you will be able to identify which object in particular I am talking about as well as I can"

béu and English are exactly the same in this respect.

* Not necessarily by using a finger ... a gesture with the head ... or even the orientation of the eyes can suffice.

..

..... Kenzuma

..

béu also has what I call an extended noun phrase. An extended noun phrase is a normal NP with either a partitive appended to the LHS, or a RC appended to the RHS.

The example below shows an extended noun phrase kenzuma with both a partitive AND an RC ...

..

SW 090.png


uya yiŋkai ofa nài tunheu-h doik-u-r-a
..... three of girl five REL townhall-DAT walk-3PL-IND-PRES

..... Three of the five girls that are walking to the townhall.

..

... The relative clause

..

The béu relative clause is pretty similar to the English relative clause. However not exactly so.

A relative clause is a clause that modifies a NP of course. I think the best way to explain how the béu RC works is to give three examples. Each example will demonstrate a subtype of RC. In each example I will reconstitute the plain clause (PC) underlying the RC by looking at the NP and the RC.

(1)

yiŋkai ofa nài doik-u-r-a
the girl five REL walk-3PL-IND-PRES

=> the five girls who are walking

NP = yiŋkai ofa : RC = nài doikura => PC = yiŋkai ofa doikura "five girls are walking" ....... notice that nài is binned.

In the above PC yiŋkai is absolutive.

(2)

bàu nài-h glá-s fy-o-r-i yiŋkai-wo
the man REL-DAT women-ERG tell-3SG-IND-PAST girl-ABOUT

=> the man to whom the woman told about the girl

NP = bàu : RC = nàih glás fyori yiŋkaiwo => PC = bàuh glás fyori yiŋkaiwo ............ notice that nài is again binned. Also -h has to find some other word to stick on to.

In the above PC bàu is dative.

(3)

gwai.a nài polg-ai-r-a fía ?ode
the islands REL sail-1PL.INC-IND-PRES between them

=> "the islands that we are sailing between"

NP = gwaia : RC = nài polgaira fía ?ode => PC = polgaira fía gwaia ...................... nài is again binned. Also ?ode is discarded. The NP must be positioned behind fía, the preposition that governs it.

In the above PC gwaia is not absolutive, also not adorned by a pila?o. Instead it exists in a prepositional phrase. For this reason, a pronoun ?ode is needed in the RC to represent the NP

..

I believe that Arabic structures its RC in a similar way to the above.

..

OK ... you should all be experts in RC's now. You just run backward the 3 NP + RC => PC processes.

..

This is discussed in greater detail in CH5.

..

... The partitive

..

A few sections back I mentioned ... the béu equivalent to "other/others/the other/another/the others".

is used where the speech participants have agreed on the population (of whatever noun category) under consideration but one of them wants to expand this population.

This expansion is a bit like "a shot in the dark", the speech participant requesting additional items usually is in the dark as to that additional items are available. Because of this, there is only one word . I mean, if the speech participant requesting additional items had an idea about what additional items were available, he could add more detail along with his request. Perhaps we would have donu meaning "another with a bell", doni meaning "another with a whistle" ... well O.K. I am being a bit facetious ... but you understand what I am getting at.

Now is used to expand the population under consideration ... to increase the scope of the conversation ... to sort of "zoom out".

Now sometimes it is necessary to "zoom in". For instance suppose you heard "three of the doctors decided to stop off at the pub on the way home" within larger narrative. After this point, these three doctors could be referred to as they. The main-protagonists/subject/topic have been reduced from eight to three. Zooming in is not a shot in the dark. The population under consideration is a known concept. The usual method is to specify the "new scope" plus the "original scope" in some sort of construction. The languages of the world all have methods for zooming in ... usually some quite simple construction, often involving a particle which has evolved from "from"/"out of". is the particle used in béu. Some examples of its use ...

ú wì moltai = all of the doctors

yè wì moltai = some of the doctore = several of those doctors = a number of those doctors

jù wì moltai = none of the doctors

tontu wì moltai dí = the majority of these doctors

a?a lú tuge wì moltai dè more = one or more of those doctors

hài wì moltai dè = many of those doctors

ima ín wì moltai dè = any two of those doctors

moltai wì bawa dí = the doctors out of these men

[ Note ... ú wì moltai = "all of the doctors" is pretty similar to ú moltai and moltai ú ... similarly these expressions with . Nothing really to worry about. You have similar flexibility in many languages (including English). Perhaps ú wì moltai stresses that no "zooming" is happening. Perhaps ú moltai is used for generic statements such as ú moltai r jini "all doctors are clever". Perhaps moltai ú is used in non-generic statements such as moltai ú ture tìa pà "all the doctors came to my home this morning" ] ..

I suppose the nearest equivalent of is "of". However has not so many functions as "of". For "belonging to", is used. For "relating to"/"connected with". is used.

[Still thinking if should be involved with "a glass of milk"/"a heart of gold"]

..

Pronouns are used in partitives in the same way as NP's.

a?a wì pài = "one of us"

ima wì onde = "two of them"

há ima wì onde glá timpura = "two of them are hitting the woman"

[ I guess English is a bit stange w.r.t. plural pronoun. I mean ... why not "*two of they are walking" {following the pattern "they are walking"} instead of "two of them are walking" ]


Two other numeratives that we haven't mentioned yet are tontu "the majority"/"most" and tonji "the minority".

ton = bit/part/section ... tontu <= ton jutu ... tonji <= ton tiji ... toŋko = to seperate ???

..

... 16 useful little words

..

jupu nobody upu everybody
juku never uku always
juda nowhere uda everywhere
jufen nothing ufen everything

..

The above are obviously eroded forms of jú pú, ú pú, ... etc. etc.

..

pu.en somebody pu.in anybody
kyu.en sometime kyu.in any time
da.en somewhere da.in anywhere
fenen something fenin anything

..

The above are obviously eroded forms of pú èn, pú ín etc. etc. They are all two syllable words, and of course as two syllable words lack tone.

The words in the first column above can be made plural by adding ... pu.in yè = some people, kyu.en yè = sometimes, etc. etc.

..

upu and ufen can be followed by uwe to show that the group is acting (or undergoing an action) collectively.

a?awe can be added to show that the group is acting (or undergoing an action) individually.

..

..... Olus

..

olus can mean "noun". It can also mean "noun phrase" (NP).

Whereas kenʒi refers to countable nouns, olus is the term used for uncountable nouns (also called mass nouns).

The expansion of olus by various modifiers is quite similar to kenʒi ...


SW 091.png

Of course numbers, iyo and hài are not appropriate (quantity slot)

Also the items in the determiner slot are a bit "iffy". For example moze dí is perhaps kài moze dí with kài elided.

..

So ... an example of an olus ...

hoŋko ima wì ?azwo pona = "two cups of warm milk"

..

A few hundred words have a dual existence ... in one guise olus in another guise kenʒi. With final vowel e u a o or i (the last one is especially common) they have a collective meaning. For example ...

..

bodi ng-o-r
birds fly-3SG-IND

=>small birds fly ................. [notice the third person singular agreement on the verb]

..

However with a change of the final vowel to ai these concepts become countable.

..

bodai lail-o-r-a
a small bird sing-3SG-IND-PRES

=> a small bird is singing

..

Which can be made plural by putting a number in front (or one of the other numeratives).

..

bodai uya lail-u-r-a
small bird three sing-3PL-IND-PRES

=> three small birds are singing

..

Here are some more of these concepts treated in this way ...

..

yinki crumpet yinkai a young unmarried woman, an attractive girl, a virgin
toti children totai a child
wazbo distance wazbai 3,680 m (the unit of distance ... the béu km or mile)
malkufa cabbages malkufai a cabbage
alha flowers alhai a flower

.. TW 793.png

..

Words derived using the suffixes mi/mai also pattern with these dual identity words. For example ... beumai = "somebody with knowledge of the béu language and/or culture" : beumi = "the entire body of people with knowledge of the béu language and/or culture"

..

Remember that all collectives take singular pronouns and, if they are A or S arguments, produce an -o- in slot 1 of the verb (as opposed to -u-).

..

There is a particle k+, that when put in front of a saidau or a kenʒi gives an olus

You hear it a lot prefixed to animal names ... like when talking about characteristics which are common to an entire species. For example ...

sadu "elephant" ... k+ sadu "the elephants" or "elephants" ... as in k+ sadu r jodo jini "the elephant is an inteligent animal"

gèu "green" .......... k+ gèu "the green ones"


k+ sadu r jodo jini
elephant-kind COP animal clever

Note ... k+ is in free variation with kài "sort/type/kind"


** Birds smaller than pidgeons are bodai. Birds that are pidgeon size and above are jwado ... jwado is kenʒi.

..

..... Saidau

..

The saidau (adjective) has two uses in béu. It can either be part of a NP or it can be a copular complement. For example ...

bàu gèu = a/the green man

bàu r gèu = a/the man is green

gèu above is a simple adjective. Adjective phrases exist as well.

An important particle that increases the degree of an adjective is sowe. For example ... gèu sowe "very green"

..

These adjectives can become nouns by froning them with ə, and kuwai.

ə gèu = a/the green one

kə gèu = a/the green ones

kuwai gèu = greenness


[ NOTE : I don't think the schwa is visually distinct enough. From now on I will use a plus sign to depict the schwa ]


+ gèu = a/the green one

k+ gèu = a/the green ones

kuwai gèu = greenness


OK ... that's better.


+ and k+ are historically derived from "one" and kài "type". Actually they are in free variation with their historical counterparts ... a bit like "either" in English can have two pronounciations. When you want to emphasize, you would of course use the phonetically heavier version.

kuwai is a word meaning property/characteristic.

Actually these 3 words are also productive with "locatives" and "genitives" as well. For example ...

+ pobomau = the one on top of the mountain

+ yó jene = the one belonging to Jane

..

TW 918.png

..

The above chart shows the main derivational pathways in béu. Only pathways 2, 3, 4 are relevent to this section.

..

Note ... + gèu sowe = "a/the very green one" ... sowe never modifies a senko.

By the way ... determiners and relative clauses can also stand by themselves, but they are unmodified when they do so. (Note to self : are you sure about this ?)

..

..... Pila?o

..

In total there are 17 cases plus the unmarked case (the absolutive case). The absolutive is not called a case in the béu linguistic tradition : instead it is called "noun base"

These 17 cases are called pila?o.

These are attached to a noun and show the relationship of that noun with respect to the rest of the sentence.

..

The word pila?o is built up from ;-

pila (v) = to place, to position, to correctly align

pila?o (n) = the positioner

..

Probably the most important case is the ergative (the 11th case). In English it is the order of the verb and the arguments that shows who is the doer and what is the "done to". Namely the A and S argument come before the verb and the O argument after ... [ English is a non-ergative language and hence the A and S argument get treated in the same way.]

In béu, to show who is the doer and what is the "done to", the suffix -s is appended to the A argument. For example ...

..

glás bàu timporI => The woman hit the man ............... (with "the man" being the O argument)

glá bàus timporI => The man hit the woman ................. (with "the man" being the A argument)

bàu tìah doikori => The man walked home ........................... (with "the man" being the S argument) ... [ béu is an ergative language and hence the O and S argument have the same form.]

..

SW 063.png

..

There is a regular relationship between preposition and affix, apart from (11) which is highly irregular, (16) which is irregular and (17) which is very slightly irregular. When suffixes they all are usually written using a single consonant. No confusion can arise as normally consonants are illicit word finally. However there is no abbreviated forms for (15) and (17). Of the 17 consonants, ? and n are not involved in these abbreviations.

..

The pila?o are either realized as either affixes or as prepositions.

Whether the pila?o appears as an suffix or a preposition depends on whether you have a N (noun) or a NP (noun phrase). If you have N the affix is used, if you have NP the preposition is used.

tiadua = beyond the house

dùa tìa yó yinkai hauʔe = beyond the house of the pretty girl

..

TW 940.png

SW 092.png

SW 093.png

SW 094.png

TW 568.png


Note on the script ... If they are realized as affixes then, in the béu script uses a sort of shorthand. That is the affix is represented as one letter.

..

Earlier we have seen that when 2 nouns come together the second one qualifies the first.

However this is only true when the words have no pila?o affixed to them. If you have two contiguous nouns suffixed by the same pila?o then they are both considered to contribute equally to the sentence roll specified. For example ...

jonos jenes solbur moze = "John and Jane drink water"

In the absence of an affixed pila?o, to show that two nouns contribute equally to a sentence (instead of the second one qualifying the first) the particle should be placed between them. For example ...

jono lé jene maumur = "John and Jane sleep"

Compare the above two examples to jono jene maumor = "Jane's John sleeps" ... that is "the John that is in a relationship with Jane, sleeps".

..

.. As parts of speech

..

pila?o of location phrases (i.e. nouns with 1 -> 8 or 15) can be considered adjectives. They must come after a noun or a verb.

pila?o of motional phrases (i.e. nouns with 13, 14, 16 or 17) can be considered adverbs. They can come in any position because it is understood that they are qualifying the verb.

pila?o phrases defining sentence rolls (i.e. nouns with 9, 10, 11 or 12) can come anywhere. They are considered clause arguments.


(Note to self : move the below to a different section)

* [ Notice that in English, you can either say ... "a bird is in the tree" or "in the tree is a bird"

In béu only jwado r ʔupaiʔe is valid ... also note that in this case jwado is not definite because it is left of the verb. That rule doesn't work with the copula. ]


jenes solbori moʒi lé ʔazwo = "Jane drank water and milk"

jonos jenes hecuri sadu ima lé ʔusfa uya = John and Jane saw two elephants and three giraffes.

This word is that is never written out in full but has its own symbol. See below ...

..

..... Maŋga

..

These are verbs. In particular maŋga = "the infinitive form of the verb" or "verbal noun". They call it "maSdar" in Arabic.

I shall call it "base verb".

In the BLT (béu linguistic tradition) the base verb is considered "dead" or "inactive". Only when a suitable "tail" is added does it become "alive" or "active".

So "finite verb" is called maŋga ?algu. The infinitive form of the verb is simply maŋga

(By the way ... ?algu = alive) ... the next Chapter is all about maŋga ?algu.

..

In béu the base verb is a noun. This is different to most languages in the world. In most languages, verbal nouns are derived from verbs. For example ... destroy => destruction : verb => noun whereas in béu. For example ... twá : "a/the meeting" => twarua : "I intend to meet".

..

English is very chaotic as to the various means it derives nouns from verbs. For example ... "discover" + "y" => the discovery ... "destroy" + "tion" => the destruction ... "run" + => a/the run. Whereas béu is as orderly as it is possible to get.

OK ... lets get started ...

solbe = "to drink"

Now the manga can amalgamate with other elements. For example ...

solbe saco = "to drink quickly" or "drinking quickly"

...and adding more elements ...

solbe moze sacois* = "to drink the water quickly" or "drinking the water quickly"

solbe moze sacois and solbe saco are examples of maŋguma (maŋga plus the augmentive -uma)

Actually a maŋguma can get as involved as a clause can. With arguments to the left of the head considered definite and arguments to the right, indefinite. But there is one difference. The ergative argument is marked by the particle rather than or -ta or -s.

so adding even more elements ...

hí jono solbe moze sacois hí jono = "John drinking the water quickly" or "for John to drink the water quickly".

This is as involved as I will go. But note that other clausal elements ( dative object, time, adverb, instrument, reason, purpose) can be added. As in normal clauses there is a tendancy to add them to the RHS.

Now the maŋguma is basically a noun ... as is kenʒi and olus. And as with kenʒi and olus more modifiers can be added.

..

SW 095.pngI don't know the béu name for this. Actually I don't know the English name for this.

The items in the determiner slot are quite rare compared to kenʒi, but they are licit. There can be nothing in the quantity slot or the ownership slot.

The "number + " modifier is quite rare. It means "so many iterations of the action". For example ...

uya wì hí pà solbe moze = (the) three times I drank water

..

Adjectives can be a bit confusing. For example saco "quick" would be used immediately after maŋga. But if not immediately after maŋga but within the maŋguma, it should take the form sacois (-is being the adjective => adverb suffix). But is it is outwith the maŋguma in the quality slot it will be saco. Got it ?

..

In the example we are using sacois "quickly" can be taken out of the heart, and placed in the senko phrase as saco. In the adjective slot of course.

In a similar vain you have a choice as to where to put a locative. A locative can be placed in the locative slot or it can be placed in maŋguma. No change of form for the locative ...

solbe moze sacois tiapi hí jono = solbe moze sacois hí jono tiapi = "John drinking the water quickly in the house" = "for John to drink the water quickly in the house".

OK ... so much for maŋguma.

..

Now we have already introduced the pila?o. The pila?o are totally compatible with kenʒi. Nearly totally compatible with olus. However only two pila?o fit in with maŋga. These are pila?o 2 and 9 ... and . When fitted to maŋga they produce an adjective and an adverb respectively.

..

We'll talk about and adverbial construction first ...

tore doikatu = "he/she came on foot" or "he/she came by walking"

tore tú doika saco = "he/she came by walking quickly"

Notice that the particle acts as it normally does and appends to the end of a single word, but stands alone to the left of a multi-word phrase.

..

And the adjectival constuction ...

..

..

bàu doika-la man walk-1SG-IND-PRES
bàu r doika-la man COPULA} walk-P2

Also appears often in conjunction with manga

The -constuction acts as an adjective. An adjective meaning "XXX-ing" at the (relevant ???) moment of speech". As with all adjectives it can either be part of a NP or it can be a copular complement. For example ...

bàu doikala = a/the walking man

bàu r doikala = a/the man is walking .... [Note ... bàu r doikala means exactly the same as bàu doikora]

differs from most other pila?o in that, with a manga, it never stands alone. For example ...

bàu doikala sacois = a/the quickly walking man .... [Note ... the affix -is is appended to saco to show it is connected to doika and not bàu] instead of *bàu là doika saco

In a -constuction, everything has the same order as a MP ... the only difference is that -la is appended to the manga and XXX is dropped. Well XXX represents the A argument and the A argument is the thing being described by the -constuction, so no need to exist inside the construction.

This -constuction can be called the present participle. The present participle has the meaning "in the process of XXXing".For example ...

doika "to walk" => doikala "in the process of walking"

kata "to cut" => katala "in the process of cutting".

When derived from a transitive verb the object can be included as well. For example katala lazde "in the process of cutting the grass".

[ Note ... bàu katala lazde "the man cutting the grass" means the same as bàu nàis katora lazde "the man who is cutting the grass" ... however the first is nearly always preferred ... well it is shorter ]

[ Also note ... pà r katala lazde means the same as (pás) katara lazde ... however the second is nearly always preferred ... well it is shorter ]

O arguments (in an equivalent active clause) can be modified by the -construction as well. For example ... lazde jwola kata "grass being cut" ... jwola kata being classed as an adjective phrase (jwòi meaning "to undergo").

..

maŋga ... as well as appearing as arguments in a clause. That is S, A, O, CS and CO, also appear as complements to auxiliary verbs.

One such auxilliary is tuma meaning "to squeaze" or "to force". [ when it means "to squeaze" it is followed by a kenʒi and is acting as a normal verb, when it means "to force" it is followed by a manga and is acting as an auxiliaryl verb ]


In these constructions, there is a very strong tendency (almost a rule) that the maŋga is on RHS of maŋguma. However if an ergative (A) argument is present, that element can instead be on RHS. Also a very strong tendency fot the O aurgment to directly follow the maŋga. There is a fairly strong tendency for the S argument to directly follow the maŋga.

So ...

Thomas forced John to hit Jane => tomos tumori timpa jene hí jono or tomos hí jono timpa jene

Thomas forced Jane to walk => tomos tumori doika jene

[By the way ... as an example of tuma being a normal verb ... tomos jwuba komo jene tumori = Thomas squeazed Jane's left buttock ]

..

Two other examples of maŋga with auxilliary verbs (why not) ...

1) ... mbe = to hold ..... lelpa = to sing, singing ..... jenes mbor lelpa bòi = Jane can sing well. [lelpai = a song ?]

2) ... glù = to depart ... timpa = to hit, hitting ... jonos glori timpa jene = John stopped hitting Jane

..

One notable use of the maŋga is emphasis, where the manga is used right next to the same word in r-form. For example ...

daw-o-r-u dàu
die-3SG-IND-FUT death

= He/she will die a death => He/she will die for sure

..

lay-o-r-i lái
live-3SG-IND-PAST life

= He/she lived a life => He/she had a full life

..

maum-a-r-i mauma
sleep-1SG-IND-PAST sleep

= I slept a sleep => I had a deep and satisfying sleep


Now maumori mauma and daw.oru dàu are strange. Normally both verbs are strictly intransitive. But here there are transitive. Seemingly sometimes béu allows

a) intransitive => transitive b) A argument => absolutive form c) O argument = maŋga

pà maumari mauma is the answer to mìn maumari mauma

.. ..

* Another way to say this is solbe saco moze**

** If saco doesn't immediately follows the manga, it must be explicitly tagged as an adverb by the -is suffix. The exact same rule as for (finite)clauses.


Note ... "I saw a man cutting the grass" is an English clause. I think Dixon analyses "the man cutting the glass" as a complement clause ??? This sees a bit strange to me. The béu equivalent .... hecari bàu katala lazde is just analyzed as Verb hecari ... Object bàu and Adjective Phrase katala lazde


..... Feŋgi

..

The feŋgi or particles are too diverse to say anything meaningful about them here. We will learn them one by one as we go though the ten chapters.

..

But just to fill out this section a bit, I will give you two sets of pronouns. One set being the pronouns in their unmarked form* and the other ... the pronouns in their ergative form**.

Here, for a transitive clause, "that which initiates the action" is called the A argument, and "that which is affected by the action" the O argument. Also, for an intransitive verb, the noun is called the S argument. It is convenient to make a distinction between all three cases. I follow RMW Dixon in using this terminology.

In most languages the S argument is marked the same way as the A argument. However in some languages the S argument is marked the same way as the O argument. These are called ergative languages. béu is one of these ergative languages. About a quarter of the world languages are ergative or partly ergative. The ergative system ...

..

bàu glá maum-u-r-i
man and woman sleep-3PL-IND-PAST

==> The man and the woman slept


bàu-s glá timp-o-r-i
man-ERG woman hit-3SG-IND-PAST

==> The man hit the woman


bàu glá-s timp-o-r-i
man woman-ERG hit-3SG-IND-PAST

==> The woman hit the man

..

Below are the béu pronouns for the S and O arguments. This form can be considered the "unmarked form".

..

SW 119.pngSW 108.png

..

mpau includes the listener in the "we", pài excludes the listener from the "we" and indicates a plural "you".

The proper way to use ?o is to use it for all non-humans. But in actually practice, some people might "elevate" the status of a family pet and call it òn or ʃì.

onde is used for "they" when they are all male, ʃide is used for "they" when they are all female, ʃinde is used for "they" when they are mixed, and ?ode is used for "they"when they are all non-human.

..

Below are the béu pronouns for the A argument ... the "ergative form".


SW 120.png

..

NOTE ... Pronouns differ from nouns in that their tones change between the ergative and the unmarked form. For a normal noun it is sufficient that -s is suffixed. For example ...

From now on I will call the ergative form the s-form, and the unmaked form the base form.

..

There is one other pronoun ... the reflexive pronoun . This is always an O argument. Notice that it is the only O argument with a high tone.

..

* In the Western Linguistic Tradition, these "forms" are called "cases". The English word case used in this sense comes from the Latin casus, which is derived from the verb cadere, "to fall", from the Proto-Indo-European root *ḱad-. The Latin word is a calque of the Greek πτῶσις, ptosis, "falling, fall". The sense is that all other cases are considered to have "fallen" away from the nominative (considered the unmarked form in Latin).

** By the way, there are 17 marked forms (cases) in béu ... the ergative being one of these.

..

... Index

  1. Introduction to Béu
  2. Béu : Chapter 1 : The Sounds
  3. Béu : Chapter 2 : The Noun
  4. Béu : Chapter 3 : The Verb
  5. Béu : Chapter 4 : Adjective
  6. Béu : Chapter 5 : Questions
  7. Béu : Chapter 6 : Derivations
  8. Béu : Chapter 7 : Way of Life 1
  9. Béu : Chapter 8 : Way of life 2
  10. Béu : Chapter 9 : Word Building
  11. Béu : Chapter 10 : Gerund Phrase
  12. Béu : Discarded Stuff
  13. A statistical explanation for the counter-factual/past-tense conflation in conditional sentences