{e'i} in usage and in the wiki

8 views
Skip to first unread message

Gleki Arxokuna

unread,
Nov 21, 2014, 3:47:32 AM11/21/14
to bpfk...@googlegroups.com
http://www.lojban.org/tiki/BPFK+Section%3A+Irrealis+Attitudinals
"Attitudinal. Used to command / dictate / instruct / order / oblige / constrain / impose an obligation."

This might be in conflict with the current usage meaning roughly "have to"

May be it's not in conflict but then explanations and more examples must be given.
Is "command to myself" really "I have to"?

I voted against this page because of {e'i} in particular too.

Jorge Llambías

unread,
Nov 21, 2014, 4:19:32 PM11/21/14
to bpfk...@googlegroups.com
On Fri, Nov 21, 2014 at 5:47 AM, Gleki Arxokuna <gleki.is...@gmail.com> wrote:
http://www.lojban.org/tiki/BPFK+Section%3A+Irrealis+Attitudinals
"Attitudinal. Used to command / dictate / instruct / order / oblige / constrain / impose an obligation."

This might be in conflict with the current usage meaning roughly "have to"

May be it's not in conflict but then explanations and more examples must be given.
Is "command to myself" really "I have to"?

I would translate for example "e'i mi cliva" into Esperanto as "mi foriru". which is a first person imperative. English doesn't strictly have first person imperatives, so it would be "I should leave".

mu'o mi'e xorxes

Alex Burka

unread,
Nov 21, 2014, 4:21:19 PM11/21/14
to bpfk...@googlegroups.com
Er, how is this not redundant to {.ei}? {.ei mi cliva} is what I've always said for that example.

mu'o mi'e la durkavore
--
You received this message because you are subscribed to the Google Groups "BPFK" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bpfk-list+...@googlegroups.com.
To post to this group, send email to bpfk...@googlegroups.com.
Visit this group at http://groups.google.com/group/bpfk-list.
For more options, visit https://groups.google.com/d/optout.

Jorge Llambías

unread,
Nov 21, 2014, 4:31:57 PM11/21/14
to bpfk...@googlegroups.com
On Fri, Nov 21, 2014 at 6:21 PM, Alex Burka <dur...@gmail.com> wrote:
Er, how is this not redundant to {.ei}? {.ei mi cliva} is what I've always said for that example.

They are similar, yes. The main difference is that with ".e'i" the speaker is the source of authority, whereas with ".ei" it's just how the speaker feels things ought to be. With first person actions the difference can be very subtle, but with second or third person it's more clear: "ei" is "this ought to happen", whereas ".e'i" is "let this happen, make this happen". "ei gusni" "there ought to be light (but there's nothing I can do about it)" "e'i gusni" "let there be light (I command that it be so)".

Gleki Arxokuna

unread,
Nov 22, 2014, 1:22:06 AM11/22/14
to bpfk...@googlegroups.com
How can "I have to go" and "I should go" be subtle? Their difference is very important.
Why losing it?


mu'o mi'e xorxes
 

--

And Rosta

unread,
Nov 22, 2014, 2:28:24 AM11/22/14
to bpfk...@googlegroups.com

English "must" is sometimes to encode,  in contrast to "have to", that the speaker is the source of the obligation. So "e'i mi cliva" would be like "I must go", while "ei mi cliva" would be like "I should go" or "I have to go". But these meanings are modal rather than illocutionary.

--And.

--

Jorge Llambías

unread,
Nov 22, 2014, 8:43:35 AM11/22/14
to bpfk...@googlegroups.com
On Sat, Nov 22, 2014 at 3:22 AM, Gleki Arxokuna <gleki.is...@gmail.com> wrote:

How can "I have to go" and "I should go" be subtle? Their difference is very important.
Why losing it?

Which difference do you think is being lost?

 .e'i is part of a series of illocutionaries e'a/e'e/e'i/e'o/e'u with the speaker's attitude towards an action to be performed (typically but not necessarily) by someone else:

.e'a: I hereby grant permission for this to take place:
.e'e: I hereby give encouragement for this to take place:
.e'i: I hereby impart instructions for this to take place: (or to use "constraint": I hereby impose a constraint for this to take place:) 
.e'o: I hereby pose a request for this to take place:
.e'u: I hereby make a suggestion for this to take place:

.ei is used to express how one feels the world ought to be:

.ei: It ought to be the case that this takes place:

So we have:

 .ei mi cliva
 It ought to be the case that I leave.
 I ought to leave.
 I should leave.

 .e'i mi cliva
 I hereby impart instructions that I leave.
 (Direct first person singular imperative not available in English, but "e'i mi'o cliva" "Let's leave".)

 [ju'a] mi bilga lo ka cliva
 [I assert that] I am under the obligation of leaving.
 I have to leave.

There are circumstances when all three can be felicitously used, even though their precise meanings are somewhat different.

Examples with "mi" as the agent are not always the best to show how UIs work, because it makes it harder to disentangle the effect on "mi" as the speaker from the effect on "mi" as the agent. With "do" we have:

 .ei do cliva
 It ought to be the case that you leave.
 You ought to leave.
 You should leave. (But not because I say so.)
 It's best if you leave.

 .e'i do cliva
 I hereby impart instructions that you leave.
 You should leave. (Because I say so. An instruction.)
 Leave.

 [ju'a] do bilga lo ka cliva
 [I assert that] You are under the obligation of leaving.
 You have to leave. (A statement, not an instruction.)

With a third person:

 .ei la .alis. cu cliva
 It ought to be the case that Alice leaves.
 Alice ought to leave.
 Alice should leave. (But not because I say so.)
 It's best if Alice leaves.

 .e'i la .alis. cu cliva
 I hereby impart instructions that Alice leaves.
 Alice should leave. (Because I say so. An instruction.)
 (Let Alice leave.)

 [ju'a] la .alis. cu bilga lo ka cliva
 [I assert that] Alice is under the obligation of leaving.
 Alice has to leave. (A statement, not an instruction.)

Gleki Arxokuna

unread,
Nov 22, 2014, 9:37:10 AM11/22/14
to bpfk...@googlegroups.com
2014-11-22 16:43 GMT+03:00 Jorge Llambías <jjlla...@gmail.com>:

On Sat, Nov 22, 2014 at 3:22 AM, Gleki Arxokuna <gleki.is...@gmail.com> wrote:

How can "I have to go" and "I should go" be subtle? Their difference is very important.
Why losing it?

Which difference do you think is being lost?

 .e'i is part of a series of illocutionaries e'a/e'e/e'i/e'o/e'u with the speaker's attitude towards an action to be performed (typically but not necessarily) by someone else:

.e'a: I hereby grant permission for this to take place:
.e'e: I hereby give encouragement for this to take place:
.e'i: I hereby impart instructions for this to take place: (or to use "constraint": I hereby impose a constraint for this to take place:) 
.e'o: I hereby pose a request for this to take place:
.e'u: I hereby make a suggestion for this to take place:

.ei is used to express how one feels the world ought to be:

.ei: It ought to be the case that this takes place:

But that's how BPFK pages repurposed ei and e'i. 
Usage tells that:
ei mi cliva - conventions/rules dictate that I leave.
e'i mi cliva - the situation forces me to leave.
ei is a UI for {sei mi te javni}.
e'i is a UI for {sei mi bilga}.

So my question is why e'i was moved from {ei} group into {e'V} group turning it into kinda imperative marker.

Jorge Llambías

unread,
Nov 22, 2014, 11:33:24 AM11/22/14
to bpfk...@googlegroups.com
On Sat, Nov 22, 2014 at 11:37 AM, Gleki Arxokuna <gleki.is...@gmail.com> wrote:

Usage tells that:
ei mi cliva - conventions/rules dictate that I leave.
e'i mi cliva - the situation forces me to leave.
ei is a UI for {sei mi te javni}.
e'i is a UI for {sei mi bilga}.

That would mean "e'i" can only be used with "mi" as agent, but I see a lot of usage of these words with sentences where "mi" is not the agent. Some random examples for ".e'i" from korpora zei sisku:

do sanga.e'i

.i mi do viska sei.e'ido terpa

.au .ei.e'ido cilre fi lo cnima'o

.i.e'ilo xekri cu kelci

.i.e'izo_cu .e nai zo_xu co'e

.uinai.e'ido toljundi mi

.i lo te pilno .e lo te xamgu cu.e'ise setca
 
.i ku'i.e'imi'o denpa

.e'iko ta mi dunda

 
So my question is why e'i was moved from {ei} group into {e'V} group turning it into kinda imperative marker.

I don't think it was moved as such. CLL doesn't even give an example with ".e'i", so it's not very clear that it wasn't in the e'V group from the start. Since you can use ".e'i" for a first person imperative anyway, it just makes sense to define it as a regular member of the e'V group.

Gleki Arxokuna

unread,
Nov 23, 2014, 4:01:51 AM11/23/14
to bpfk...@googlegroups.com
2014-11-22 19:33 GMT+03:00 Jorge Llambías <jjlla...@gmail.com>:


On Sat, Nov 22, 2014 at 11:37 AM, Gleki Arxokuna <gleki.is...@gmail.com> wrote:

Usage tells that:
ei mi cliva - conventions/rules dictate that I leave.
e'i mi cliva - the situation forces me to leave.
ei is a UI for {sei mi te javni}.
e'i is a UI for {sei mi bilga}.

That would mean "e'i" can only be used with "mi" as agent

oops, sorry. the two my definitions of {e'i} contradict each other.
Let me try once again:
ei mi cliva - conventions/rules dictate that I leave.
e'i mi cliva - the situation forces me to leave.

ei = sei [lo nu no'a cu] se javni
e'i = sei [lo nu no'a cu] se bapli [be lo nau tcini]

by [lo nu no'a] I mean the event of the bridi in which the sei-clause is inserted.


, but I see a lot of usage of these words with sentences where "mi" is not the agent. Some random examples for ".e'i" from korpora zei sisku:

Do you think they reflect the meaning that you suggest for {e'i}?
E.g. this is how I understand them (recurring examples are not translated):


do sanga .e'i  = It has to be so that you sing.

.i mi do viska sei .e'i do terpa = ?

.au .ei .e'i do cilre fi lo cnima'o = It should and has to be so that you learn about cnima'o, and I wish it was so.

.i .e'i lo xekri cu kelci

.i .e'i zo_cu .e nai zo_xu co'e

.uinai .e'i do toljundi mi

.i lo te pilno .e lo te xamgu cu .e'i se setca
 
.i ku'i .e'i mi'o denpa = We have to wait

.e'i ko ta mi dunda = It has to be so that you wait so wait!

I can agree with And that "to have to" might not be an illocutionary thing but rather a modal but then how can these examples be understood otherwise?
 



 
 
So my question is why e'i was moved from {ei} group into {e'V} group turning it into kinda imperative marker.

I don't think it was moved as such. CLL doesn't even give an example with ".e'i", so it's not very clear that it wasn't in the e'V group from the start. Since you can use ".e'i" for a first person imperative anyway, it just makes sense to define it as a regular member of the e'V group.

mu'o mi'e xorxes

Jorge Llambías

unread,
Nov 23, 2014, 9:27:39 AM11/23/14
to bpfk...@googlegroups.com

On Sun, Nov 23, 2014 at 6:01 AM, Gleki Arxokuna <gleki.is...@gmail.com> wrote:
oops, sorry. the two my definitions of {e'i} contradict each other.
Let me try once again:
ei mi cliva - conventions/rules dictate that I leave.
e'i mi cliva - the situation forces me to leave.

ei = sei [lo nu no'a cu] se javni
e'i = sei [lo nu no'a cu] se bapli [be lo nau tcini]

by [lo nu no'a] I mean the event of the bridi in which the sei-clause is inserted.

Do you really want "nau" or "lo tcini be lo nu no'a"? 

 
, but I see a lot of usage of these words with sentences where "mi" is not the agent. Some random examples for ".e'i" from korpora zei sisku:

Do you think they reflect the meaning that you suggest for {e'i}?

For the most part, yes.
 
E.g. this is how I understand them (recurring examples are not translated):

do sanga ,e'i  = It has to be so that you sing.

"Sing!"

.i mi do viska sei .e'i do terpa = ?

"I see you - be afraid (of that)."
 
.au .ei .e'i do cilre fi lo cnima'o = It should and has to be so that you learn about cnima'o, and I wish it was so.

A mix of "I wish you would learn about attitudinals", "you ought to learn about attitudinals" and "do learn about attitudinals".
  
.i ku'i .e'i mi'o denpa = We have to wait

"But let's wait."
 
.e'i ko ta mi dunda = It has to be so that you wait so wait!

"Give that to me!" 

I can agree with And that "to have to" might not be an illocutionary thing but rather a modal but then how can these examples be understood otherwise?

As imperatives. The modal is "na ka'e na" or "bi'ai".
Reply all
Reply to author
Forward
0 new messages