xorlo and the sumti gerku

33 views
Skip to first unread message

Tomislav Ostojich

unread,
Oct 2, 2017, 4:06:21 AM10/2/17
to lojban-b...@googlegroups.com
I’ve yet to find a good systematic treatment of xorlo despite searching, so I am asking someone to do the obvious: take the sumti “gerku” and apply every possible xorlo to it, followed by a brief English translation and explanation.

guskant

unread,
Oct 2, 2017, 8:51:51 AM10/2/17
to Lojban Beginners


Le lundi 2 octobre 2017 08:06:21 UTC, Tomislav Ostojich a écrit :
I’ve yet to find a good systematic treatment of xorlo despite searching, so I am asking someone to do the obvious: take the sumti “gerku” and apply every possible xorlo to it, followed by a brief English translation and explanation.


I'm sorry that you might have considered my commentary as a bad systematic treatment of xorlo.

Even though it is bad, it includes a systematic table of meaning of {lo} with numbers in Section 3.2.1. By applying M=2, N=5 and broda={gerku}, we obtain the following examples:

re lo mu gerku :
two (which are dogs distributively) of dogs that are five

re loi mu gerku :
two (which are dogs distributively) of groups that consists of dogs that are five

re lo'i mu gerku :
two (which are dogs distributively) of sets that consists of dogs that are five

pi re lo mu gerku :
Quantity 0.2 of a part of one of dogs that are five
(It is not a dog but a part of a dog.)

pi re loi mu gerku :
Quantity 0.2 of a part of one of groups that consists of dogs that are five
(It is one dog.)

pi re lo'i mu gerku :
Quantity 0.2 of a part (subset) of one of sets that consists of dogs that are five
(It is a set of one dog.)

mu'o 

guskant

unread,
Oct 2, 2017, 8:42:12 PM10/2/17
to Lojban Beginners


Le lundi 2 octobre 2017 12:51:51 UTC, guskant a écrit :



re loi mu gerku :
two (which are dogs distributively) of groups that consists of dogs that are five

re lo'i mu gerku :
two (which are dogs distributively) of sets that consists of dogs that are five


Sorry, the sentences in the parentheses must be corrected as follows:

re loi mu gerku :
two (which are groups distributively) of groups that consists of dogs that are five

re lo'i mu gerku :
two (which are sets distributively) of sets that consists of dogs that are five

mu'o

MorphemeAddict

unread,
Oct 2, 2017, 9:03:13 PM10/2/17
to lojban-b...@googlegroups.com
pi re loi mu gerku :
Quantity 0.2 of a part of one of groups that consists of dogs that are five
(It is one dog.)

Why does this not also mean an arbitrary group of parts of the group of five dogs, such as all their heads or legs? 

stevo

--
You received this message because you are subscribed to the Google Groups "Lojban Beginners" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban-beginners+unsubscribe@googlegroups.com.
To post to this group, send email to lojban-beginners@googlegroups.com.
Visit this group at https://groups.google.com/group/lojban-beginners.
For more options, visit https://groups.google.com/d/optout.

guskant

unread,
Oct 2, 2017, 11:18:20 PM10/2/17
to Lojban Beginners


Le mardi 3 octobre 2017 01:03:13 UTC, stevo a écrit :
pi re loi mu gerku :
Quantity 0.2 of a part of one of groups that consists of dogs that are five
(It is one dog.)

Why does this not also mean an arbitrary group of parts of the group of five dogs, such as all their heads or legs? 

stevo

 
Depending on the interpretation of combination of {gerku} and {si'e}, it can mean heads or tails of dogs, but I don't think that interpretation reflects a usual meaning of {gerku} and {si'e}. 

{pi PA SUMTI} is defined as {lo pi PA si'e be pa me SUMTI}. According to usual meaning of {si'e}, the referant of x1 of {pi PA si'e} should have the same property as the referent of x2 of {pi PA si'e}, though that interpretation is not officially declared. As shown in the example {pi re lo mu gerku}, some selbrisni like {gerku} do not fit the usual meaning of {si'e} because a part of dog is not a dog in usual meaning. 

In the case of {pi re loi mu gerku}, which is developed as {lo pi re si'e be pa me lo gunma be lo mu gerku}, x2 of {pi re si'e} refers to a collective group of the referent of {gerku} in the meaning of xorlo, and then the referent of x2 has the property of {gerku}. Therefore, according to the usual meaning of {si'e}, x1 of {pi re si'e} should have the property of {gerku}. However, according to the usual meaning of {gerku}, a head or a tail of {gerku} does not have the property of {gerku}. In order to satisfy the usual meaning of both, x1 of {pi re si'e} should be one dog.

mu'o
Reply all
Reply to author
Forward
0 new messages