Lesson 18 pabimoi seltadni Possession nunponse 18.1 On po and po'e me zo po .e zo po'e We will cover two more members of selma'o GOI at this time, "po" and "po'e". These two elements deal with an aspect related to identification: possession. In Lojban, we recognize two distinction types of possession: - alienable; - inalienable. We'll define these in a moment. The definition of possession is probably too culturally oriented to make a clean line between possession and other relationships, or even among various types of possession. We must allow for possession to be taken in the broadest sense - the set of all things, states, properties, etc. which may pertain to a person or be identified with a person in any way, at any time. This obviously goes far beyond legal ownership. Yet, we must allow for people who want a narrower definition. Let us look at some possible 'possessions' under a fairly broad definition. The following are typical things that might be 'possessed'. - a book; - a house; - an arm; - a brain; - a heart; - a book you have written; - an address; - a name; - an idea; - a kiss; - love. You can legally own a book. You can be holding a book which you may or may not legally own. The book may be located near you, such that it can identified by the fact that it is near you. In all these cases, the book may be said to be yours. Some of these relationships are definitely 'possession' relationships, some are borderline, and some are 'possessions' only in the weakest sense. The weakest form can be demonstrated by putting a book in front of each (of several) person(s) which is not legally owned by that person. Clearly you can then tell George to pick up 'his' book, and he will know what you mean. Does he possess the book (before he picks it up)? [I - The instructor may want to quickly act out these types of possession.] All of these ways of possessing a book are called 'alienable possession'. Alienable possession refers to any object, state, or property which you possess, or which pertains to you, for which that possession is not inherently permanent. In Lojban, something may be considered a 'possession' if it pertains or is relevant to you in any way; eventually usage may set some standards. The possession is 'alienable' if there are conditions in which that possession or relevance ceases. If there is some relationship under all conditions, the possession is 'inalienable'. If it can be taken away from you such that it is not yours, it is alienable. A book which you legally own is yours, but if you sell it or give it away, it is no longer 'your' book. If it is stolen, it is your book only in the sense that you might eventually get it back. The same is true for almost any other type of possession in the physical or legal sense. You live in a house. Whether you legally own it or not, it can be said to be 'your house'. 'Your address' fits the same constraints, even though it is not physical, and cannot be legally owned. If you move away from your house, it is no longer 'yours'. On the other hand, your arm is 'inalienably' yours. If John were to chop of his arm and hand it to me, it would still be "John's arm". It could be burned or otherwise destroyed, but we would still think of it as John's arm. Even after John is dead, it is still 'his arm'. No legal, physical, or emotional means can take it away in this alienable sense. In general, body parts are inalienable, but there are some fuzzy aspects. If John dies, and his heart is transplanted into another person, it is still John's heart in one sense; in this sense it is an inalienable possession. It is also the heart of the new 'owner', as well. John's blood remains inalienably his in one sense, even when donated. But when mixed with other blood, that identity is lost. The blood is no longer 'his', though individual blood cells might be thought of in this way. Most intangibles are inalienable. Your ideas remain yours in some sense, though they may not be recognized by some people as yours. Your actions are yours. If you kiss another person, it is still your kiss. Emotions are inalienable; this is how we can talk about 'giving away love' and still having it. If brains are ever transplanted, there will be questions as to whose brain it is after the transplant. The answer will probably depend on how people think about it. This reveals something important about possession: the division between alienable and inalienable is distinct, but it is culturally based. Since the lines between types of possession are culturally defined, we must be able to talk of any possession as if it were alienable or inalienable, or to refer to it as holding a relationship that is not a true 'possessive'. The distinctions, while sometimes cultural, are important enough to rate assigning separate words to each type. We use "po" to attach an identification to a sumti of an 'owner' for which the possession is alienable or ambiguous. We use "po'e" to identify ownership as inalienable. The speaker is the one who draws the line as to whether the possession is alienable or not. We also have "pe", discussed in the next section, for relationships of relevance that we cannot, or do not want to, classify as "possessions". Thus, at least in our culture, we might say: le birka [ku] po'e la djan. "The arm of John"; "John's arm" le cukta [ku] po mi "The book of me", "my book" Note that we did not translate the latter as 'the book of mine'. "The book of me" in English could be a book about me or a book which I own. The translation is accurate; either type of relationship to the book could be described as a possession of a sort, and thus as "le cukta [ku] po mi" in Lojban. "po" is a weak form of possession; it identifies only by saying that the relative sumti in some way considered to be a possession by the relativized sumti. "po'e" makes the stronger identification of inalienable possession. 18.2 On pe me zo pe We have solved nearly all problems associated with "po'u"; two important ones remain to haunt us. How can we express a restriction which involves time and location tenses? How can we restrict using a sumti without implying any of the cultural connotations of possession? Let us look at an example of the former from English. Time and space associations are among a set of loose associations that are sometimes hard to distinguish from possession in English. If you take five people and put a book on the floor in front of each, in English you might tell someone to get "John's book" to refer to the book in front of John. Yet John's only association with the book is one of location. In Lojban, we could use the tagged sumti "va la djan." to locate the book, but how do we attach this sumti as a restriction: - we could use a bridi with "poi" as stated above: "poi jibni la djan." This is fine for simple tenses but will prove unwieldy for more complex ones; - we would prefer not to use "po" or "po'e", attempting to reserve these specifically for relationships that can be described as possession; - we cannot use "po'u"; "va la djan." is not another identity for the book; it doesn't identify the book at all, except in a relative sense. The latter can be made clear by expanding the complete sumti to its corresponding form with "du": ?le cukta po'u va la djan. becomes: le cukta poi ke'a du va la djan. Remember that "va la djan." is a sumti that is implicitly not part of the normal place structure of du. If we say: le cukta cu du va la djan. we mean an ellipsized: le cukta cu du zo'e va la djan. We are holding that the (ellipsized) identification is valid in a noteworthy way when "near John" is applicable to the bridi relationship, suggesting perhaps that the relationship may not be valid when not near John. A separate cmavo in GOI is used for expressing this sort of generalized restrictive relationship. The word is "pe", and it joins a relative sumti onto another sumti, just as all other members of GOI do. Unlike all of the other GOI members that we have seen so far, "pe" doesn't make any implication about the nature of the restriction. Time- or location-tagged modifier sumti are among the most frequent restrictions attached with "pe"; others are permitted, especially when you wish to avoid using a possessive in a way which might erroneously imply ownership. Specific usage is up to the speaker; the definition and significance of 'possession' is probably too culturally-oriented to attempt to dictate what is attached with "po", and what is attached with "pe". Just remember that "po'u" is not applicable unless you are satisfied that both sumti are in themselves alternate designations for the referent. Let us look at some examples of how we can use "pe" to attach a restriction to something by associating it with a time or a place. For the relative sumti, we use a time- or location-tagged modifier sumti. The modifier may be either completely expressed, or it may be elliptical: le cukta [ku] "the book which is located at the table" "the book on the table" le cukta [ku] {pe (pu [ku]) [ge'u]} "the book which is associated with being before something (unspecified)" "the former book" Some interesting variations on tense inflections can be obtained with "pe": mi [cu] pu tcidu le cukta [ku] mi pu tcidu le cukta "I previously-eat the food." "I ate the food." mi pe pu [ku] [ge'u] [cu] tcidu le cukta [ku] mi pe pu tcidu le cukta "The me which is associated with the previous-to-something eats the food." "I ate the food." The first sentence emphasizes the eating as being in the past; this places the event of eating in the past, and emphasizes the former nature of the relationships involved in eating. This is because of the close similarity between the tense inflection and a tense-tagged modifier sumti. The second sentence places the "me" who is doing the eating in the past. In addition to placing emphasis on that particular sumti, it heightens the implication that the un-modified sumti may not be true: the 'current me' may not be eating. Neither of the two sentences claims anything about the present. By emphasizing the past when tense omission is the norm, they make implications about the present. The examples given may seem stilted to English speakers. There are a few English constructions that use tensed 'possessions' in a way that seems natural. "Tomorrow's leaders" could translate as: "le ba lidne" Similarly, "the wonders of tomorrow" could be expressed as: "le se manci pe ba" Besides attaching tagged sumti as relative phrases restricting a sumti, "pe" is also used whenever we want to restrict a sumti by attaching a relative sumti which is closely associated with the relativized sumti, but which is not truly 'possessed' by the relative sumti. Many phrases that are expressed as possessives in English should use "pe" rather than "po". You thus use "le cukta pe la djan." to refer to a book that may incidentally be near John, or which he is holding, rather than the stronger possessive "le cukta po la djan." 18.3 An Example Dialog mupli nuncasnu Following is a dialog which uses some of the relative clause and phrase features discussed in this sublesson. The instructor may choose to have students act it out, as practice in using relative clauses. The dialog is given first without translation, so that you can first attempt to understand it on your own. .alis.: xu do djuno le se zvati pe la fred. doi rik. rik.: la fred. ki'a .alis.: mi djica la fred. poi lojbo tadni ku ku le nu cpedu zo'e rik.: mi .ia kakne le nu te cpedu .alis.: na go'i .i .ua la fred. ca klama .i co'o doi rik. rik.: co'o. .alis. .alis.: doi fred. .i ko denpa mi fred.: mi denpa .alis.: doi fred. .i xu do djuno le jbena detri po'e la rik. fred.: .ia go'i .i li rere pi'eci .alis.: ki'e doi fred. .i mi djica le nu le pendo po la rik. cu jikca penmi la rik. ca le detri .i ko mipri le nu penmi kei la rik. fred.: .ai 18.4 Translation Of The Example Dialog mupli nuncasnu xe fanva .alis.: xu do djuno le se zvati pe la fred. doi rik. Is-it-true-that you know the at-place relating-to Fred, O Rick? rik.: la fred. ki'a Fred who? .alis.: mi djica la fred. poi lojbo tadni ku ku le nu cpedu zo'e I want the Fred who Lojban studies, in order to do the act-of requesting (something unspecified). rik.: mi .ia kakne le nu te cpedu I (certainly) can do the act-of receiving-requests. .alis.: na go'i .i .ua la fred. ca klama .i co'o doi rik. It is not so. (Discovery) Fred is now coming. 'Bye, O Rick. rik.: co'o. .alis. 'Bye, Alice. .alis.: doi fred. .i ko denpa mi O Fred, wait-for me. fred.: mi denpa I wait. .alis.: doi fred. .i xu do djuno le jbena detri po'e la rik. O Fred, is-it-true-that you know the born date of Rick? fred.: .ia go'i .i li rere pi'eci (Certainly) It is so. 22 [day] 3 [month]. .alis.: ki'e doi fred. .i mi djica le nu le pendo po la rik. cu penmi la rik. ca le detri .i ko mipri le nu jikca penmi kei la rik. Thanks, O Fred. I want the act-of the friend(s) of Rick meeting Rick on the date. Keep-secret the act-of socially meeting from Rick. fred.: .ai (Aye, Intention). 18.5 On da me zo da We have been using "zo'e", the 'elliptical sumti', as a place holder in bridi for several lessons. We usually translate this as "something (unspecified)". The ambiguity in this English translation is heightened by the vague nature of the English word "something". The word is strictly a place- holder and not truly a pro-sumti because "zo'e" used in other sumti places does not refer to the same sumti value. "zo'e" has a specific meaning: there is a particular something that the speaker has in mind to fill in the place held by "zo'e", but for some reason the speaker is not specifying it. This might be to de- emphasize the place or possibly because the speaker can't or chooses not to describe what she/he has in mind. There is another meaning to the English word "something", one which is legitimately a pronoun. This meaning is what is often called the 'existential' pronoun. Logicians use it frequently in such sentences as: "There exists something x, such that x is a book.", or more briefly "Something is a book." The existential pronoun "something" does not imply that the speaker has anything particular in mind to fill the value; it merely states that the universe contains one or more possible replacements for the variable, and says nothing about the possible replacement(s) other than that it(they) is(are) book(s). Lojban has a separate series of words to cover the existential meaning of "something". These are the da-series of selma'o KOhA: "da", "de", and "di", which can be translated as "some x1", "some x2", and "some x3", which should not be confused with the sumti place-labels. Logicians use the letter designations to indicate that existential pronouns are variables; the specific term used for members of the da-series is 'bound variables'. (Lojban does provide for an indefinite additional number of bound variables beyond these 3 by explicitly using subscripts; this is described in a later lesson.) As implied above, "da" is a true pro-sumti. Once used in a reference, it holds its value indefinitely. If we say: mi klama zo'e zo'e "I go to somewhere from somewhere." we might be referring to separate places. If we say: mi klama da da we refer to "something" that is both origin and destination. On the other hand, the "zo'e" version could be referring to the same place - it is not forbidden. The same is true if two different existentials are use: mi klama da de does not imply that "de" refers to something different than "da", although it explicitly allows the possibility. Note that this meaning of "something" also includes "someone"; there is no implication whether the variable is an object or a person in: "For some x, x is covered with cloth." In English, however, we cannot help but infer constraints on 'x' in statements like "There exists some x, such that x is a book." (more colloquially: "There is something (x1) which is a book." or even "Something is a book.") It doesn't seem likely to that we can use "someone" in any of these phrasings. The Lojban equivalent explicitly denotes that no such constraints may be assumed: "da" is not restricted from being a person in "da cukta". The next section will discuss how to constrain existential variables. There is one significant semantic difference between the da-series and other members of selma'o KOhA that justifies use of separate lexemes. The 'transformational' difference is in the way that quantification (e.g. "pa da" vs. "pa do") is interpreted. In review, in a description such as: "pa le ci mensi" ("one of the three sisters"), the second quantifier ("ci") indicates the number in the set being described as sisters; the first quantifier merely selects from that set. Remember also that cmavo of KOhA by themselves imply nothing directly about number; they may be either singular or plural. When other members of selma'o KOhA are quantified, it does not resolve the question of singular vs. plural. In "pa do", "do" is presumed to be a known referent, and the sumti is literally translated as "one of you" and not "you (singular)"; there could be more than one of you, but only one is being referred to. This is same as the first quantifier in the 'sisters example. Unlike KOhA and "le" descriptions, when a member of the da-series is first used, it has no particular referent; it is not yet 'bound' to anything specific, but could refer to anything or anybody in the universe, and it might refer to any number of things or persons. Attaching a quantifier constrains an unbound "da" or "de" or "di", binding it to some subset of the entire universe. Such a quantifier thus specifies the number in the subset. "pa da" is singular; it means "one something x1", and not "one of the somethings x1". Alternatively, you can think of "pa da" meaning "one of all 'somethings' in the universe", which, while stilted, recalls the unbound nature of "da". The usage of the da-series in logic makes the binding of "da" an enumeration of the set being referred to. We can use this property of the da-series to quantify a variety of things without otherwise describing them. For example, to enumerate whether "mi" is singular or plural, we might say: mi goi pa da "I/we who am/are one something" "I (singular)" mi goi su'ore da "I/we who am/are at least two somethings" "I (plural)" Quantified da-series cmavo can also be used in a regular bridi in ways that will seem unusual to English speakers. If we want to say: "The table has four legs." in Lojban, we do not want to use "ponse" to translate "has". The relationship isn't really one of possession in the normal sense of the word, and the use of "ponse" invites debate on whether that is the relationship that applies. The brivla "tuple" is sufficient, and emphasizes that "legs" are what we are talking about, not possession: le jubme se tuple vo da "The table is be-legged by four somethings." or its inverse: vo da tuple le jubme "Four somethings are-legs of the table." Note that once "da" has been enumerated, or constrained in any way, it is now 'bound'. Quantification must then be treated as for any other specific sumti - it selects a smaller subset from the now constrained subset of the universe. Thus: le jubme se tuple vo da .i pa da crino "The table is be-legged by four somethings. One of them (the legs) is green." 18.6 On da poi me lo'u da poi le'u The "da" usage of "something" is not common in English; there is usually some constraint that we wish to put on the variable - there is some concept we have in mind. "da" is used a lot more in Lojban due to its special quantificational property just described. We need, however, to be able to express "someone", thus implying a person, as opposed to the vaguer "some x1". How might we do this? We could express such a specific pronoun with "ko'a"-series variables and "goi": ?ko'a goi le prenu This is valid grammatically; it presumes, though, that we know who "le prenu" is. It states that "ko'a" will be used to refer to the one being described as a person. This is a constraint on "ko'a", but not really the one we want. What we really want to do is define a restriction on "da" or one of its relatives. We can do this straightforwardly using "poi": da poi prenu cu klama "Something which is a person comes." "Someone comes." We can now say that a table has four green legs: le jubme se tuple vo da poi crino "The table is be-legged by four somethings such that they are green." Note that this again binds "da" and later quantifications select a subset from the four that are referred to by "da". 18.7 An Activity selzukte If the class is not reasonably comfortable with the vocabulary and grammar presented so far, the instructor may want to skip this activity for now, possibly going back to it in a few lessons. It can easily come to involve considerable spontaneous usage, for which the class may not be ready. On the other hand, it might be worth trying simply to see how close the class is to being ready to try spontaneous conversation. The instructor should practice this exercise with a partner ahead of time, planning out where to put the various objects, and writing out some sample sentences ahead of time which use the various grammatical points raised. For this activity, you need several objects, some of which are very similar or identical. These should be placed around the room. The instructor should then use sentences using relative clauses of various types in order to ask students to get particular items. The following examples use a series of sentences involving books: ko cpacu le cukta po'u la baibl. ko dunda le cukta le ninmu poi pritu la djan. ko cpacu le cukta po la carlet. ko cpacu le cukta pe vi le loldi ko jgari le cukta le pritu birka po'e do 18.8 Incidental Relative Clauses and Phrases *** Recall that we defined "po'u" as reflecting one variety of "du" bridi, specifically one in which the right side restrictively identifies an unknown qleft side: "la djil. du mi" "The one called Jill is identical to me." "la djil. po'u mi" "The one called Jill who is identical to me ..." In these examples, Jill is presumably not known to the listener, but certainly the listener can tell who is speaking (We will ignore the complications of the vagueness of "mi" and presume that there is a single speaker who is the obvious referent of "mi".) We said earlier that "po'u" is not entirely symmetric in the same way "du" is, in that the reversal of the second expression no longer makes sense as a restrictive phrase: "mi du la djil." "I am identical to the one called Jill." "mi po'u la djil." "I who am restricted to the one called Jill ..." Given our assumption of the one plausible interpretation of "mi", the identification as Jill is unnecessary as a restriction. Instead of restricting, the phrase really is merely providing additional information about "mi". This type of relative phrase is called a incidental (or non-restrictive) relative phrase. These occur in English, as well as in most (but not all) other languages. In Lojban, we want to make clear when a relative clause or phrase is providing useful identification or restriction, and when it is merely conveying incidental information. The semantic distinction can be seen in an American English example of restrictive vs. incidental clauses: "The man who came to dinner is a friend." "The man, who came to dinner, is a friend." The first sentence has a restrictive clause; the man is identified as the one who came to dinner (as opposed to some other man). The second phrase is incidental. To an American English listener, the impression is given that the speaker believes that the listener already knows who the man is, and is just mentioning in passing that he came to dinner. (The emphasis on American English is due to the standard British usage, which has commas around both restrictive and incidental clauses, so that there is no way to tell the difference.) A incidental relative clause is logically identical to two separate sentences. In the above example, these are: "The man came to dinner. The man is a friend." Because of this equation, a sentence with a incidental relative clause is false if either the main predication or the relative predication is false. Since a restrictive relative clause serves only to identify the relativized phrase, it doesn't really affect the truth value of the main predication of the sentence, so long as it adequately restricts the relativized phrase. Assuming that the distinction is now clear, how do you distinguish between restrictive and incidental relative clauses and phrases. The answer is simple; the cmavo used for non-restrictive (incidental) clauses and phrases starts with 'n', instead of 'p'. Thus the above example of a incidental phrase should have been: "mi du la djil." "I am identified as the one called Jill." My name is Jill. "mi no'u la djil." "I, who am incidentally the one called Jill, ..." The relative clause examples may be expressed as: "le nanmu poi klama le vanci sanmi cu pendo" "The man who came to dinner is a friend."(American) "The man, who came to dinner, is a friend."(British) "le nanmu noi klama le vanci sanmi cu pendo" "The man, who came to dinner, is a friend." which is equivalent to: "le nanmu cu klama le vanci sanmi .i ra pendo" "The man came to dinner. The man is a friend." There is a subtle semantic difference between the combined incidental sentence and the two separate sentences. While they are logically equivalent, the incidental expression subordinates the one sentence to the other. It is less important, perhaps only incidental information; in any case, the speaker is for some reason de-emphasizing it. With two separate sentences, there is no reason to presume that the two bridi are not equally important. In fact, in the order presented, the intended subordinate clause might be taken as the more important point, since in general a more important statement is expressed before a less important one (unless there is some other reason for the ordering such as causality or time ordering). The distinction between restrictive and incidental clauses is one of the trickiest points of English grammar, partly due to the similarity between their expressions. In learning Lojban, the vital importance of relative clauses and phrases makes understanding the difference even more important, yet Lojban's distinctions make the two concepts much easier to understand and distinguish. Do you agree? Of the cmavo for relative clauses and phrases we have thus far discussed, only "poi", "pe", and "po'u" have incidental equivalents; as you would expect, they are "noi", "ne", and "no'u"). The other two, "po", and "po'e", are used in particular for restricting in ways that are both necessary and common. Their corresponding incidental equivalents are not especially significant to expression, and do not justify the use of three cmavo to express them. Instead, you can use "noi" followed by an appropriate bridi to express these as incidental clauses. Examples "la pasifik. noi xamsi cu stici la .iunaitydsteits." "The Pacific, which is incidentally an ocean, i s to the west of the United States." "le stici be la .iunaited. steits. xamsi no'u la pasifik. cu barda" "The west-of-the-United-States-ocean, incidentally identified as the Pacific, is big." "lemi re kanla noi blanu cu kalri" "My two eyes, which are blue, are open." "ledo kanla poi blanu cu kalri .i ledo kanla poi bunre cu ganlo" "Your eye which is blue is open. Your eye which is brown is closed."