{"id":2793,"date":"2012-04-24T05:45:50","date_gmt":"2012-04-24T04:45:50","guid":{"rendered":"http:\/\/emilkirkegaard.dk\/en\/?p=2793"},"modified":"2012-04-24T05:47:04","modified_gmt":"2012-04-24T04:47:04","slug":"thoughts-about-an-introduction-to-language-fromkin-et-al","status":"publish","type":"post","link":"https:\/\/emilkirkegaard.dk\/en\/2012\/04\/thoughts-about-an-introduction-to-language-fromkin-et-al\/","title":{"rendered":"Thoughts about: An Introduction to Language (Fromkin et al)"},"content":{"rendered":"

Victoria Fromkin, Robert Rodman, Nina Hyams – An Introduction to Language<\/a><\/p>\n

I thought i better read a linguistics textbook before i start studying it formally. Who wud want to look like a noob? ;)<\/p>\n

I have not read any other textbook on this subject, but i think it was a fairly typical okish textbook. Many of the faults with it are mentioned below in this long ‘review’.<\/p>\n

Chapter 1<\/h3>\n

–<\/p>\n

In the Renaissance a new middle class emerged who wanted their children
\nto speak the dialect of the \u201cupper\u201d classes. This desire led to the publication of
\nmany prescriptive grammars. In 1762 Bishop Robert Lowth wrote A Short Intro-
\nduction to English Grammar with Critical Notes. Lowth prescribed a number
\nof new rules for English, many of them influenced by his personal taste. Before
\nthe publication of his grammar, practically everyone\u2014upper-class, middle-class,
\nand lower-class\u2014said I don\u2019t have none and You was wrong about that. Lowth,
\nhowever, decided that \u201ctwo negatives make a positive\u201d and therefore one should
\nsay I don\u2019t have any; and that even when you is singular it should be followed by
\nthe plural were. Many of these prescriptive rules were based on Latin grammar
\nand made little sense for English. Because Lowth was influential and because
\nthe rising new class wanted to speak \u201cproperly,\u201d many of these new rules were
\nlegislated into English grammar, at least for the prestige dialect\u2014that variety of
\nthe language spoken by people in positions of power.
\nThe view that dialects that regularly use double negatives are inferior can-
\nnot be justified if one looks at the standard dialects of other languages in the
\nworld. Romance languages, for example, use double negatives, as the following
\nexamples from French and Italian show:<\/p>\n

French: Je ne veux parler avec personne.
\nI not want speak with no-one.<\/p>\n

Italian: Non voglio parlare con nessuno.
\nnot I-want speak with no-one.<\/p>\n

English translation: \u201cI don\u2019t want to speak with anyone.\u201d<\/p>\n

Lowth seems to have done a good thing with his reasoning, which was obviously inspired from math: multiplying two negatives does give a positive (-1*-1=+1). The reason is logic, altho predicate logic which wasnt invented at his time (i.e., in the 1700s).<\/p>\n

Formalizing the negro english sentence “I don\u2019t have none” yields something like this: \u00ac\u2203x\u00acHix — it is not the case that there is something such that i dont have it. which is equivalent with: \u2200xHix — For any thing, i have that thing [i.e. i have everything]. Ofc, it may seem that with this remark im begging the question, but the formalization wud be closer to the natural language which is always a good thing. Im not begging the question with that remark.<\/p>\n

Furthermore, his rule made the language simpler as one no longer had to needlessly inflect the frase “anyone” into its negative form “no one”. Simpler languages are better if they have the same expressive power. Doing away with a needless inflection is good per definition makes the language simpler without losing expressive power.<\/p>\n

He was wrong about the thing with “you was”. It wud have been nice if it had stayed that way. Then english cud have begun moving towards the simplicity of verb conjugation in scandinavian.<\/p>\n

—<\/p>\n

When we say in later chapters that a sentence is grammatical we mean that it
\nconforms to the rules of the mental grammar (as described by the linguist); when
\nwe say that it is ungrammatical, we mean it deviates from the rules in some way.
\nIf, however, we posit a rule for English that does not agree with your intuitions
\nas a speaker, then the grammar we are describing differs in some way from the
\nmental grammar that represents your linguistic competence; that is, your lan-
\nguage is not the one described. No language or variety of a language (called a
\ndialect) is superior to any other in a linguistic sense. Every grammar is equally
\ncomplex, logical, and capable of producing an infinite set of sentences to express
\nany thought. If something can be expressed in one language or one dialect, it
\ncan be expressed in any other language or dialect. It might involve different
\nmeans and different words, but it can be expressed. We will have more to say
\nabout dialects in chapter 10. This is true as well for languages of technologically
\nunderdeveloped cultures. The grammars of these languages are not primitive or
\nill formed in any way. They have all the richness and complexity of the gram-
\nmars of languages spoken in technologically advanced cultures.<\/p>\n

Stupid relativism. Of course some dialects and languages are superior to others! The awful german grammar system is much inferior to the simpler scandinavian systems or the english system. More difficult it is to say which of those systems are superior to which. English has gotten rid of grammatical gender (good!) but returns pointless verb conjugations (bad!) in scandinavian there are grammatical genders (bad, but only 2 not 3 as in german) but much less pointless verb conjugation (good!).<\/p>\n

Why do the authors write this relativism nonsense? They dislike language puritanists:<\/p>\n

Today our bookstores are populated with books by language purists attempt-
\ning to \u201csave the English language.\u201d They criticize those who use\u00a0 enormity to
\nmean \u201cenormous\u201d instead of \u201cmonstrously evil.\u201d But languages change in the
\ncourse of time and words change meaning. Language change is a natural pro-
\ncess, as we discuss in chapter 11. Over time enormity was used more and more
\nin the media to mean \u201cenormous,\u201d and we predict that now that President
\nBarack Obama has used it that way (in his victory speech of November 4, 2008),
\nthat usage will gain acceptance. Still, the \u201csaviors\u201d of the English language will
\nnever disappear. They will continue to blame television, the schools, and even
\nthe National Council of Teachers of English for failing to preserve the standard
\nlanguage, and are likely to continue to dis (oops, we mean disparage) anyone
\nwho suggests that African American English (AAE)4 and other dialects are via-
\nble, complete languages.
\nIn truth, human languages are without exception fully expressive, complete,
\nand logical, as much as they were two hundred or two thousand years ago.
\nHopefully (another frowned-upon usage), this book will convince you that all
\nlanguages and dialects are rule-governed, whether spoken by rich or poor, pow-
\nerful or weak, learned or illiterate. Grammars and usages of particular groups
\nin society may be dominant for social and political reasons, but from a linguistic
\n(scientific) perspective they are neither superior nor inferior to the grammars
\nand usages of less prestigious members of society.<\/p>\n

They are right to be annoyed at the purists, they are wrong to completely abandon prescriptive grammar because of it. (Baby, bathtub)<\/p>\n

–<\/p>\n

To hold that animals communicate by systems qualitatively different from
\nhuman language systems is not to claim human superiority. Humans are not
\ninferior to the one-celled amoeba because they cannot reproduce by splitting
\nin two; they are just different sexually. They are not inferior to hunting dogs,
\nwhose sense of smell is far better than that of human animals. As we will discuss
\nin the next chapter, the human language ability is rooted in the human brain,
\njust as the communication systems of other species are determined by their bio-
\nlogical structure. All the studies of animal communication systems, including
\nthose of primates, provide evidence for Descartes\u2019 distinction between other ani-
\nmal communication systems and the linguistic creative ability possessed by the
\nhuman animal.<\/p>\n

More relativism. So, humans are not inferior to dogs with regards to smelling.. they are just.. olfactory challenged?<\/p>\n

With thing with reproduction is harder. Asexual and (bi)sexual reproduction both have some advantages and disadvantages. Cellular division wud obviously not work for humans (we are too complex), but asexual reproduction might work somewhat. We get to try it out soon when we start cloning people. Im looking forward to when we start digging up the graves of past geniuses to make a clone of them i.e., harvest some DNA and insert it into an egg, and put that egg into a woman.<\/p>\n

–<\/p>\n

In our understanding of the world we are certainly not \u201cat the mercy of what-
\never language we speak,\u201d as Sapir suggested. However, we may ask whether the
\nlanguage we speak influences our cognition in some way. In the domain of color
\ncategorization, for example, it has been shown that if a language lacks a word
\nfor red, say, then it\u2019s harder for speakers to reidentify red objects. In other words,
\nhaving a label seems to make it easier to store or access information in memory.
\nSimilarly, experiments show that Russian speakers are better at discriminating
\nlight blue (goluboy) and dark blue (siniy) objects than English speakers, whose
\nlanguage does not make a lexical distinction between these categories. These
\nresults show that words can influence simple perceptual tasks in the domain
\nof color discrimination. Upon reflection, this may not be a surprising finding.
\nColors exist on a continuum, and the way we segment into \u201cdifferent\u201d colors
\nhappens at arbitrary points along this spectrum.<\/strong> Because there is no physical
\nmotivation for these divisions, this may be the kind of situation where language
\ncould show an effect.<\/p>\n

But this is simply not true. The segmentations are not at all arbitrary. It is strange that the authors claim this as they just reviewed information form a language that segments colors into two categories: light and dark colors. These are not arbitrary categories. I learned about this from Lakoff’s Women, Fire, Dangerous Things<\/em> (which is hosted somewhere on my site), but see also: http:\/\/en.wikipedia.org\/wiki\/Linguistic_relativity_and_the_color_naming_debate<\/a>.<\/p>\n

–<\/p>\n

Chapter 2<\/h3>\n

Additional evidence regarding hemispheric specialization is drawn from Japa-
\nnese readers. The Japanese language has two main writing systems. One system,
\nkana, is based on the sound system of the language; each symbol corresponds to
\na syllable. The other system, kanji, is ideographic; each symbol corresponds to
\na word. (More about this in chapter 12 on writing systems.) Kanji is not based
\non the sounds of the language. Japanese people with left-hemisphere damage
\nare impaired in their ability to read kana, whereas people with right-hemisphere
\ndamage are impaired in their ability to read kanji. Also, experiments with unim-
\npaired Japanese readers show that the right hemisphere is better and faster than
\nthe left hemisphere at reading kanji, and vice versa.<\/p>\n

This is pretty cool! Even better, it fits with the data from the last book i read:<\/p>\n

Visual memory is not normally tested in intelligence tests. There have been four studies of the
\nvisual memory of the Japanese, the results of which are summarized in Table 10.7. Row 1
\ngives a Japanese IQ of 107 for 5-10-year-olds on the MFFT calculated from error scores com-
\npared with an American sample numbering 2,676. The MFFT consists of the presentation of
\ndrawings of a series of objects, e.g., a boat, hen, etc. that have to be matched to an identical
\ndrawing among several that are closely similar. The task entails the memorization of the de-
\ntails of the drawings in order to find the perfect match. Performance on the task correlates
\n0.38 with the performance scale of the WISC (Plomin and Buss, 1973), so that it is a weak
\ntest of visualization ability and general intelligence as well as a test of visual memory. Row 2
\ngives a visual memory IQ of 105 for ethnic\u00a0 Japanese Americans compared with American
\nEuropeans on two tests of visual memory consisting of the presentation of 20 objects for 25
\nseconds and then removed, and the task was to remember and rearrange their positions. Row 3
\nshows a visual memory IQ of 110 obtained by comparing a sample of Japanese high school
\nand university students with a sample of 52 European students at University College, Dublin.
\nRow 4 shows a visual memory IQ of 113 for the visual reproduction subtests of the Wechsler
\nMemory Scale-Revised obtained from the Japanese standardization of the test compared with
\nthe American standardization sample. The test involves the drawing from memory of geomet-
\nric designs presented for 10 seconds. The authors suggest that the explanation for the Japanese
\nsuperiority may be that Japanese children learn kanji, the Japanese idiographic script, and this
\ndevelops visual memory capacity. However, this hypothesis was apparently disproved by the
\nFlaherty and Connolly study (1996) whose results are shown in row 2. Some of the ethnic
\nJapanese American participants had a knowledge of kanji, while others did not, and there was
\nno difference in visual memory between those who knew and those who did not know kanji,
\ndisproving the theory that the advantage of East Asians on visualization tasks arises from their
\npractice on visualizing idiographic scripts. (Richard Lynn, Race differences in intelligence, p. <\/em>94)<\/p>\n

It fits. Why else wud those people choose a very visual language instead of a more sound (i.e. verbal) focused one? Tests also show that east asians are worse at verbal tasks. This makes perfectly sense with their writing system.<\/p>\n

–<\/p>\n

Chapter 3<\/h3>\n

In the foregoing dialogue, Humpty Dumpty is well aware that the prefix un-
\nmeans \u201cnot,\u201d as further shown in the following pairs of words:
\nA —————– B
\ndesirable —— undesirable
\nlikely ———- unlikely
\ninspired ——- uninspired
\nhappy ——— unhappy
\ndeveloped—– undeveloped
\nsophisticated – unsophisticated<\/p>\n

Thousands of English adjectives begin with un-. If we assume that the most
\nbasic unit of meaning is the word, what do we say about parts of words like
\nun-, which has a fixed meaning? In all the words in the B column, un- means
\nthe same thing\u2014\u201cnot.\u201d Undesirable means \u201cnot desirable,\u201d unlikely means \u201cnot
\nlikely,\u201d and so on. All the words in column B consist of at least two meaningful
\nunits: un + desirable, un + likely, un + inspired, and so on.<\/p>\n

The authors are again wrong. The un<\/em> prefix does not mean “not” in these examples! An undesirable person is more than just someone that isnt desirable, it is someone who is, well, positively undesirable; that one wants to avoid. Similarly for likely<\/em>+unlikely<\/em>. When one says that something is unlikely, one is saying more than just saying that it is not likely. One is saying that it has a low probability of happening. The difference here is that the event cud be neither likely or unlikely, i.e. having a probability around .5 (or whatever, depends on context). An unhappy person is someone who is sad or depressed, not just someone who isnt happy. A neutral person is neither happy or unhappy. An example of a word where the un<\/em> prefix has the simple meaning of negation, is something like unmarried<\/em> which really only does mean “not married”. The un<\/em> prefix in many if not all of their examples has the function of reversing the quality in question, not negating it.<\/p>\n

I have pointed this out before, but it was in a forum post on FRDB where i am now banned and therefore cannot search using the built-in search tool.<\/p>\n

Chapter 4<\/h3>\n

Whether a verb takes a complement or not depends on the properties of the
\nverb. For example, the verb find is a transitive verb. A transitive verb requires an
\nNP complement (direct object), as in The boy found the ball, but not *The boy
\nfound, or *The boy found in the house. Some verbs like eat are optionally tran-
\nsitive. John ate and John ate a sandwich are both grammatical.
\nVerbs select different kinds of complements. For example, verbs like put <\/em>and
\ngive <\/em>take both an NP and a PP complement, but cannot occur with either alone:<\/p>\n

Sam put the milk in the refrigerator.
\n*Sam put the milk.
\nRobert gave the film to his client.
\n*Robert gave to his client.<\/p>\n

Sleep is an intransitive verb; it cannot take an NP complement.
\nMichael slept.
\n*Michael slept a fish.<\/p>\n

What about: “Sam puts out.” (see meaning #6<\/a>) That lacks a NP and is grammatical. And how about: “Robert gave a talk.” (see meaning #2<\/a>) That lacks a PP and is grammatical. It seems that the authors shud have chosen some better example verbs.<\/p>\n

–<\/p>\n

Chapter 5<\/h3>\n

For most sentences it does not make sense to say that they are always true
\nor always false. Rather, they are true or false in a given situation, as we pre-
\nviously saw with\u00a0 Jack swims. But a restricted number of sentences are indeed
\nalways true regardless of the circumstances. They are called\u00a0 tautologies. (The
\nterm analytic is also used for such sentences.) Examples of tautologies are sen-
\ntences like Circles are round or A person who is single is not married. Their
\ntruth is guaranteed solely by the meaning of their parts and the way they are
\nput together. Similarly, some sentences are always false. These are called contra-
\ndictions. Examples of contradictions are sentences like Circles are square or A
\nbachelor is married.<\/p>\n

Not entirely correct. Analytic sentences are noncontingent sentences, not just noncontingetly true sentences.<\/p>\n

http:\/\/plato.stanford.edu\/entries\/analytic-synthetic\/<\/a><\/p>\n

Later on they write:<\/p>\n

The following sentences are either tautologies (analytic), contradictions, or
\nsituationally true or false.<\/p>\n

Indicating that they think analytic<\/em> refers only to noncontingetly true propositions\/sentences. Also, they shud perhaps have studied some more filosofy, so that they wudn’t have to rely on the homemade term situationally true<\/em> when there already exists a standard term for this, namely contingently true<\/em>.<\/p>\n

–<\/p>\n

Much of what we know is deduced from what people say alongside our obser-
\nvations of the world. As we can deduce from the quotation, Sherlock Holmes
\ntook deduction to the ultimate degree. Often, deductions can be made based on
\nlanguage alone.<\/p>\n

Sadly, the authors engage in the common practice of referring to what Sherlock Holmes did as “deduction”. It wasn’t. It was mostly abduction aka. inference to the best explanation.<\/p>\n

http:\/\/plato.stanford.edu\/entries\/abduction\/<\/a><\/p>\n

–<\/p>\n

Generally, entailment goes only in one direction. So while the sentence Jack
\nswims beautifully entails Jack swims, the reverse is not true. Knowing merely that
\nJack swims is true does not necessitate the truth of Jack swims beautifully. Jack
\ncould be a poor swimmer. On the other hand, negating both sentences reverses
\nthe entailment. Jack doesn\u2019t swim entails Jack doesn\u2019t swim beautifully.<\/p>\n

They are not negating it properly. They are using what i before called short-form negation. Compare:<\/p>\n

“Jack doesn\u2019t swim” (\u2203!x)x=j\u2227\u00acSj
\nwith
\n“It is not the case that Jack swims” \u00ac(\u2203!x)x=j\u2227Sj<\/p>\n

These two do not mean the same, strictly speaking. And the distinction does sometimes matter. The one entails that Jack exists and the second does not. This matters when one is talking about sentences such as “The current king of France is bald”.\u00a0 I have explained this before.<\/a><\/p>\n

–<\/p>\n

The notion of entailment can be used to reveal knowledge that we have about
\nother meaning relations. For example, omitting tautologies and contradictions,
\ntwo sentences are\u00a0 synonymous (or paraphrases) if they are both true or both
\nfalse with respect to the same situations. Sentences like Jack put off the meeting
\nand Jack postponed the meeting are synonymous, because when one is true the
\nother must be true; and when one is false the other must also be false. We can
\ndescribe this pattern in a more concise way by using the notion of entailment:
\nTwo sentences are synonymous if they entail each other.<\/p>\n

The authors conflate ‘meaning the same’ with ‘having the same truth-value’. These are not the same. Some sentences always have the same truth-value (they belong to the same equivalence class) but do not mean the same. Examples are e.g.:<\/p>\n

“Canada is north of the US”
\nand
\n“The US is south of Canada”<\/p>\n

These two don’t mean the same, but they belong to the same equivalence class. The relation among the entities is reversed in the other sentence i.e. “… is north of …” and “… is south of …” do not mean the same. They mean the opposite of each other.<\/p>\n

See Swartz and Bradley (1979:35ff)<\/a> for more examples and a more thoro discussion.<\/p>\n

–<\/p>\n

The semantic theory of sentence meaning that we just sketched is not the
\nonly possible one, and it is also incomplete, as shown by the paradoxical sen-
\ntence This sentence is false. The sentence cannot be true, else it\u2019s false; it cannot
\nbe false, else it\u2019s true. Therefore it has no truth value, though it certainly has
\nmeaning. This notwithstanding, compositional truth-conditional semantics has
\nproven to be an extremely powerful and useful tool for investigating the seman-
\ntic properties of natural languages.<\/p>\n

Obviously, i’m not going to let this one fly! :) Things are not nearly as simple as they write. I will just point to my friend’s, Benjamin Burgis, recent dissertation (ph.d.)<\/a> about the liar paradox and other related problems.<\/p>\n

One point tho. Note the authors strange inference from to “Therefore, it has no truth value”.<\/p>\n

–<\/p>\n

In the previous sections we saw that semantic rules compute sentence meaning
\ncompositionally based on the meanings of words and the syntactic structure that
\ncontains them. There are, however, interesting cases in which compositionality
\nbreaks down, either because there is a problem with words or with the semantic
\nrules. If one or more words in a sentence do not have a meaning, then obviously
\nwe will not be able to compute a meaning for the entire sentence.<\/strong> Moreover,
\neven if the individual words have meaning but cannot be combined together as
\nrequired by the syntactic structure and related semantic rules, we will also not
\nget to a meaning. We refer to these situations as semantic anomaly. Alternatively,
\nit might require a lot of creativity and imagination to derive a meaning. This is
\nwhat happens in metaphors. Finally, some expressions\u2014called idioms\u2014have a
\nfixed meaning, that is, a meaning that is not compositional. Applying composi-
\ntional rules to idioms gives rise to funny or inappropriate meanings.<\/p>\n

A bit of clarification is needed here. They are right if they mean the word is used in the sentence. They are wrong if they mean the word is mentioned in the sentence. The unclear frasing “in a sentence” won’t do here. See http:\/\/plato.stanford.edu\/entries\/quotation\/#2.2<\/a><\/p>\n

–<\/p>\n

The semantic properties of words determine what other words they can be com-
\nbined with. A sentence widely used by linguists that we encountered in chapter
\n4 illustrates this fact:<\/p>\n

Colorless green ideas sleep furiously.<\/p>\n

The sentence obeys all the syntactic rules of English. The subject is\u00a0 colorless
\ngreen ideas and the predicate is sleep furiously. It has the same syntactic struc-
\nture as the sentence<\/p>\n

Dark green leaves rustle furiously.<\/p>\n

but there is obviously something semantically wrong with the sentence. The
\nmeaning of\u00a0 colorless\u00a0 includes the semantic feature \u201cwithout color,\u201d but it is
\ncombined with the adjective green, which has the feature \u201cgreen in color.\u201d How
\ncan something be both \u201cwithout color\u201d and \u201cgreen in color\u201d? Other semantic
\nviolations occur in the sentence. Such sentences are semantically anomalous.<\/p>\n

The authors seem to be saying that all sentences that involves contradictions are semantically anomalous. But that is not true, if by that they mean that such sentences are meaningless. Self-contradictory sentences are meaningless alright. Otherwise, their negations (which are necessarily true) wud be meaningless too. A grammatically correct placed negation can never make a sentence meaningful or meaningless.<\/p>\n

I have discussed this before. See this essay<\/a>, and this post (by the good doctor Burgis) and the comments section below<\/a>.<\/p>\n

The authors however do mention later that:<\/p>\n

The well-known colorless green ideas sleep furiously is semantically
\nanomalous because ideas (colorless or not) are not animate.<\/p>\n

So, i’m not sure what they think. Perhaps they think that the chomsky is anomalous for both reasons, i.e. 1) that it is self-contradictory, and 2) it involves a category error with the verb sleep<\/em> and the subject ideas<\/em>.<\/p>\n

–<\/p>\n

Another part of the meaning of the words baby <\/em>and child <\/em>is that they are
\n\u201cyoung.\u201d (We will continue to indicate words by using italics and semantic fea-
\ntures by double quotes.) The word father has the properties \u201cmale\u201d and \u201cadult\u201d
\nas do uncle and bachelor.<\/p>\n

(I have restored the authors italicization in the above quote)<\/p>\n

First, it bothers me when authors want to put a given word in quotation marks but then include something that doesn’t belong in there with it, typically a comma or a dot. Very annoying!<\/p>\n

Second, they are wrong about these semantic features. The word father<\/em> has the features “parent” and “male”. It has no feature about adulthood altho that it is often the case. There is nothing semantically strange or anomalous about calling a person who is 15 years old a father if he has a child. Similar things hold about their other example uncle<\/em>.<\/p>\n

–<\/p>\n

Generally, the count\/mass distinction corresponds to the difference between
\ndiscrete objects and homogeneous substances. But it would be incorrect to say
\nthat this distinction is grounded in human perception, because different lan-
\nguages may treat the same object differently. For example, in English the words
\nhair, furniture, and spaghetti are mass nouns. We say Some hair is curly, Much
\nfurniture is poorly made, John loves spaghetti. In Italian, however, these words
\nare count nouns, as illustrated in the following sentences:<\/p>\n

Ivano ha mangiato molti spaghetti ieri sera.
\nIvano ate many spaghettis last evening.
\nPiero ha comprato un mobile.
\nPiero bought a furniture.
\nLuisella ha pettinato i suoi capelli.
\nLuisella combed her hairs.<\/p>\n

We would have to assume a radical form of linguistic determinism (remem-
\nber the Sapir-Whorf hypothesis from chapter 1) to say that Italian and English
\nspeakers have different perceptions of hair, furniture, and spaghetti. It is more
\nreasonable to assume that languages can differ to some extent in the semantic
\nfeatures they assign to words with the same referent, somewhat independently
\nof the way they conceptualize that referent. Even within a particular language
\nwe can have different words\u2014count and mass\u2014to describe the same object or
\nsubstance. For example, in English we have shoes (count) and footwear (mass),
\ncoins (count) and change (mass).<\/p>\n

But what about a nonperfect correlation? The data mentioned above does not disprove the existence of a such thing. It wud be interesting to do a cross-language study to see if there was a correlation. I wud be very surprised if there was no such correlation. I will bet money that something like this is the case: The more discrete an entity is, the higher the chance that the thing will be a countable noun. It is not surprising that their examples involves things that almost always but not always come in bundles. But i’d wager that no language has car<\/em> as a noncountable noun. The entity is too discrete for that to make sense. Likely, i’d be surprised if any language had water as a countable noun. Generally, words for fluids are probably always (or nearly so) noncountable nouns. Even if the words for the entities that these fluids are made of are countable nouns e.g. a molecule<\/em>.<\/p>\n

–<\/p>\n

In all languages, the reference of certain words and expressions relies entirely
\non the situational context of the utterance, and can only be understood in light
\nof these circumstances. This aspect of pragmatics is called deixis (pronounced
\n\u201cdike-sis\u201d). Pronouns are deictic. Their reference (or lack of same) is ultimately
\ncontext dependent.
\nExpressions such as<\/p>\n

this person
\nthat man
\nthese women
\nthose children<\/p>\n

are also deictic, because they require situational information for the listener to
\nmake a referential connection and understand what is meant. These examples
\nillustrate person deixis. They also show that the demonstrative articles like this
\nand that are deictic.
\nWe also have\u00a0 time deixis and place deixis. The following examples are all
\ndeictic expressions of time:<\/p>\n

now then tomorrow
\nthis time that time seven days ago
\ntwo weeks from now last week next April
\n…<\/p>\n

In filosofy, these are called indexicals<\/a>.<\/em> Or so i thought, apparently, there is some difference according to Wikipedia. Deixis<\/a> seems to be a bit broader.<\/p>\n

–<\/p>\n

Implicatures are different than entailments. An entailment cannot be can-
\ncelled; it is logically necessary. Implicatures are also different than presupposi-
\ntions. They are the possible consequences of utterances in their context, whereas
\npresuppositions are situations that must exist for utterances to be appropriate in
\ncontext, in other words, to obey Grice\u2019s Maxims. Further world knowledge may
\ncancel an implicature, but the utterances that led to it remain sensible and well-
\nformed, whereas further world knowledge that negates a presupposition\u2014oh,
\nthe team didn\u2019t lose after all\u2014renders the entire utterance inappropriate and in
\nviolation of Grice\u2019s Maxims.<\/p>\n

To be fair, they only talked about deductive inferences or entailment before. But some entailment maybe be ‘cancelled’ by further information or premises<\/em> as they are called in logic. Logics where new information can make an inference worse or better are called non-monotonic<\/a>.<\/p>\n

–<\/p>\n

Chapter 6<\/h3>\n

Throughout several centuries English scholars have advocated spelling
\nreform. George Bernard Shaw complained that spelling was so inconsistent that
\nfish could be spelled ghoti\u2014gh as in tough, o as in women, and ti as in nation.
\nNonetheless, spelling reformers failed to change our spelling habits, and it took
\nphoneticians to invent an alphabet that absolutely guaranteed a one sound\u2013one
\nsymbol correspondence. There could be no other way to study the sounds of all
\nhuman languages scientifically.<\/p>\n

It’s not their fault tho. Blame the politicians. As i have repeatedly shown, there are various good ways to reform english spelling. In fact, i’ve begun working on my own ultra minimalistic reform proposal. More on that later. :)<\/p>\n

–<\/p>\n

The sounds of all languages fall into two classes: consonants and vowels. Con-
\nsonants are produced with some restriction or closure in the vocal tract that
\nimpedes the flow of air from the lungs. In phonetics, the terms consonant and
\nvowel refer to types of sounds, not to the letters that represent them. In speaking
\nof the alphabet, we may call \u201ca\u201d a vowel and \u201cc\u201d a consonant, but that means
\nonly that we use the letter \u201ca\u201d to represent vowel sounds and the letter \u201cc\u201d to
\nrepresent consonant sounds.<\/p>\n

Indeed. I recall that when i invented Lyddansk (my danish reform proposal) i had to make this distinction. I called them vowel-letters and consonant-letters (translated).<\/p>\n

–<\/p>\n

5.\u00a0 The following are all English words written in a broad phonetic transcrip-
\ntion (thus omitting details such as nasalization and aspiration). Write the
\nwords using normal English orthography.
\na. [hit]
\nb. [strok]
\nc. [fez]
\nd. [ton]
\ne. [boni]
\nf. [skrim]
\ng. [frut]
\nh. [prit\u0283\u0259r]
\ni. [krak]
\nj. [baks]
\nk. [\u03b8\u00e6\u014bks]
\nl. [w\u025bnzde]
\nm. [kr\u0254ld]
\nn. [kant\u0283i\u025bnt\u0283\u0259s]
\no. [parl\u0259m\u025bnt\u00e6ri\u0259n]
\np. [kw\u0259b\u025bk]
\nq. [pits\u0259]
\nr. [b\u0259rak obam\u0259]
\ns. [d\u0292\u0254n m\u0259ken]
\nt. [tu \u03b8a\u028az\u0259nd \u00e6nd et]<\/p>\n

I really, really dislike their strange choice of fonetical symbols. They don’t correspond to major dictionaries online nor the OED. Especially confusing is using \/e\/ for both \/e\/ and \/e\u026a\/ as in eight<\/em>, which they write as \/et\/ instead of the normal \/e\u026at\/ found in pretty much all dictionaries (example: 1<\/a>, 2<\/a>, and the OED gives the same pronunciation).<\/p>\n

To those that are wondering, here is what i think the correct answers are:<\/p>\n

a. [hit] hit<\/em>
\nb. [strok] stroke<\/em> but their symbolism is confusing, they use \/o\/ to mean IPA \/\u0259\u028a\/
\nc. [fez] face<\/em>? which shud be \/fe\u026as\/
\nd. [ton] it is tempting to guess ton<\/em> until one thinks of their strange use of \/o\/ to mean \/\u0259\u028a\/, the correct word must be tone<\/em> \/t\u0259\u028an\/
\ne. [boni] bunny<\/em> is tempting, but it seems to be boney<\/em> \/b\u0259\u028ani\/
\nf. [skrim] scrim<\/em>
\ng. [frut] froot<\/em>, they fail to indicate that the vowel is long i.e. \/fru:t\/
\nh. [prit\u0283\u0259r] preacher<\/em>
\ni. [krak] crack<\/em>
\nj. [baks] backs <\/em>is tempting, but it appears to be barks<\/em> i.e. \/b\u0251\u02d0ks\/
\nk. [\u03b8\u00e6\u014bks] thanks<\/em>
\nl. [w\u025bnzde] another strange one, i think it is wednesday<\/em> i.e. \/w\u0292nzde\u026a\/
\nm. [kr\u0254ld] crawled<\/em>
\nn. [kant\u0283i\u025bnt\u0283\u0259s] conscientious<\/em>? i.e. \/k\u0252n\u0283\u026a\u02c8\u025bn\u0283\u0259s\/
\no. [parl\u0259m\u025bnt\u00e6ri\u0259n] parliamentarian <\/em>
\np. [kw\u0259b\u025bk] Quebec<\/em>
\nq. [pits\u0259] pizza<\/em>
\nr. [b\u0259rak obam\u0259] Barack Obama<\/em>
\ns. [d\u0292\u0254n m\u0259ken] John McCain<\/em>
\nt. [tu \u03b8a\u028az\u0259nd \u00e6nd et] two thousand and eight<\/em>, with eight<\/em> which shud be \/e\u026at\/.<\/p>\n

In general, their introduction to fonetics is bad when it disagrees with pretty much all dictionaries. Learn fonetics somewhere else. I learned it from Wikipedia and using lots of dictionaries.<\/p>\n

–<\/p>\n

Chapter 7<\/h3>\n

Nothing interesting to note here.<\/p>\n

Chapter 8<\/h3>\n

Some time after the age of one, the child begins to repeatedly use the same string
\nof sounds to mean the same thing. At this stage children realize that sounds are
\nrelated to meanings. They have produced their first true words. The age of the
\nchild when this occurs varies and has nothing to do with the child\u2019s intelligence.
\n(It is reported that Einstein did not start to speak until he was three or four
\nyears old.)<\/p>\n

It saddens me to see that a textbook with a chapter about children and learning spread this myth! It is not that hard to google it and discover it to be an urban myth. See: http:\/\/www.learninginfo.org\/einstein-learning-disability.htm<\/a><\/p>\n

–<\/p>\n

[b\u0259rt]\u00a0 \u201c(Big) Bird\u201d<\/p>\n

Another annoying detail with their chosen fonetical symbols is that they fail to distinguish between schwa \/\u0259\/ which is an unstressed vowel, and the similar sounding but potentially stressed vowel \/\u025c\/. Again, they don’t use the same standards as used by dictionaries, which is annoying! But see: http:\/\/en.wikipedia.org\/wiki\/Schwa<\/a> and http:\/\/en.wikipedia.org\/wiki\/Mid-central_vowel<\/a><\/p>\n

–<\/p>\n

1.\u00a0 Hans hat ein Buch gekauft. \u201cHans has a book bought.\u201d
\n2.\u00a0 Hans kauft ein Buch. \u201cHans is buying a book.\u201d<\/p>\n

I don’t get it. How can a linguistics textbook get the translation wrong? The correct translation of (2) is “Hans buys a book.”.<\/p>\n

–<\/p>\n

Another experimental technique, called the naming task, asks the subject to
\nread aloud a printed word. (A variant of the naming task is also used in stud-
\nies of people with aphasia, who are asked to name the object shown in a pic-
\nture.) Subjects read irregularly spelled words like dough and steak just slightly
\nmore slowly than regularly spelled words like doe and stake, but still faster than
\ninvented strings like cluff. This suggests that people can do two different things
\nin the naming task. They can look for the string in their mental lexicon, and if
\nthey find it (i.e., if it is a real word), they can pronounce the stored phonologi-
\ncal representation for it. They can also \u201csound it out,\u201d using their knowledge
\nof how certain letters or letter sequences (e.g., \u201cgh,\u201d \u201coe\u201d) are most commonly
\npronounced. The latter is obviously the only way to come up with a pronuncia-
\ntion for a nonexisting word.
\nThe fact that irregularly spelled words are read more slowly than regularly
\nspelled real words suggests that the mind \u201cnotices\u201d the irregularity. This may be
\nbecause the brain is trying to do two tasks\u2014lexical look-up and sounding out
\nthe word\u2014in parallel in order to perform naming as fast as possible. When the
\ntwo approaches yield inconsistent results, a conflict arises that takes some time
\nto resolve.<\/p>\n

\u00a0This is very interesting! I didn’t know that badly spelled words were read more slowly. That’s good news, or bad news, depending. :P It is good in that i may now have another argument for spelling reform: it makes people more efficient readers. It is also testable between populations+languages. Everything else equal, are people that read a well-spelled language faster readers than people that read a horribly spelled language (like english and danish)? That’s an interesting question actually. It sounds sufficiently simple and obvius that someone must have done the study. As for the bad news part, if they are right, it means i’m being inefficient becus i’m reading in a bad language. Worse, the entire world is being inefficient becus of its ‘choice’ of world language (i.e. english).<\/p>\n

Chapter 9<\/h3>\n

Some systems draw on formal logic for semantic representations. You put up
\nthe switch would be represented in a function\/argument form, which is its logi-
\ncal form:<\/p>\n

PUT UP (YOU, THE SWITCH)<\/p>\n

where PUT UP is a \u201ctwo-place predicate,\u201d in the jargon of logicians, and the
\narguments are YOU and THE SWITCH. The lexicon indicates the appropriate
\nrelationships between the arguments of the predicate PUT UP.<\/p>\n

I really, really dislike the term argument<\/em> when used to mean the thing that one puts into functions or predicates. It is really a very, very bad choice of words for the context (logic). argument<\/em> already has a rather precise meaning in that context. I prefer the term variable<\/em> but there is another and better term that i prefer more, but i can’t seem to recall it right now.<\/p>\n

–<\/p>\n

\u00a0A keyword as general as bird may return far more information than could be
\nread in ten lifetimes if a thorough search of the Web occurs. (A search on the
\nday of this writing produced 200 million hits, compared to 122 million four
\nyears prior.) […]<\/p>\n

I re-did the search. 1,100 million hits.<\/p>\n

–<\/p>\n

\u00a0Chapter 10<\/h3>\n

It is not always easy to decide whether the differences between two speech
\ncommunities reflect two dialects or two languages. Sometimes this rule-of-
\nthumb definition is used: When dialects become mutually unintelligible\u2014when
\nthe speakers of one dialect group can no longer understand the speakers of
\nanother dialect group\u2014these dialects become different languages.
\nHowever, this rule of thumb does not always jibe with how languages are
\nofficially recognized, which is determined by political and social considerations.
\nFor example, Danes speaking Danish and Norwegians speaking Norwegian and
\nSwedes speaking Swedish can converse with each other. Nevertheless, Danish
\nand Norwegian and Swedish are considered separate languages because they are
\nspoken in separate countries and because there are regular differences in their
\ngrammars. Similarly, Hindi and Urdu are mutually intelligible \u201clanguages\u201d spo-
\nken in Pakistan and India, although the differences between them are not much
\ngreater than those between the English spoken in America and the English spo-
\nken in Australia.<\/p>\n

Not citing any sources for such claims is bad. The mutual intelligibility is not that high between the scandinavian languages<\/a>. It is much higher for written text between norwegian (bokm\u00e5l) and danish. Etc. See the Wikipedia link.<\/p>\n

–<\/p>\n

English is the most widely spoken language in the world (as a first or second<\/strong>
\n language).<\/strong> It is the national language of several countries, including the United
\nStates, large parts of Canada, the British Isles, Australia, and New Zealand. For
\nmany years it was the official language in countries that were once colonies of
\nBritain, including India, Nigeria, Ghana, Kenya, and the other \u201canglophone\u201d
\ncountries of Africa. There are many other phonological differences in the vari-
\nous dialects of English used around the globe.<\/p>\n

This is certainly false. Look at Wikipedia<\/a>. Mandarin is the most spoken native language. English is probably the most spoken non-native language.
\nETA: But then later they write<\/p>\n

The Sino-Tibetan family includes Mandarin, the most populous language in <\/strong>
\nthe world<\/strong>, spoken by more than one billion Chinese. This family also includes
\nall of the Chinese languages, as well as Burmese and Tibetan.<\/p>\n

So, i don’t know what they think.<\/p>\n

–<\/p>\n

Even though every language is a composite of dialects, many people talk and
\nthink about a language as if it were a well-defined fixed system with various
\ndialects diverging from this norm. This is false, although it is a falsehood that is
\nwidespread. One writer of books on language accused the editors of Webster\u2019s
\nThird New International Dictionary, published in 1961, of confusing \u201cto the
\npoint of obliteration the older distinction between standard, substandard, collo-
\nquial, vulgar, and slang,\u201d attributing to them the view that \u201cgood and bad, right
\nand wrong, correct and incorrect no longer exist.\u201d In the next section we argue
\nthat such criticisms are ill founded.<\/p>\n

It’s time for the authors to again say negative things about language standardization, and promote a very relativistic view of languages and dialects. I will defend my views against their criticisms of such views.<\/p>\n

I don’t know about a ‘fixed’ system, if they meant unchanging system, then i ofc don’t agree that there is any unchanging system of standard english (or standard danish etc.). However, there is a kind of danish that is the most standard. It may be a good idea to speak as normal a version of a language as possible, becus this makes it the easiest for the listeners to understand what one is saying. The general idea is to avoid things that are peculiar to a small minority of the speakers of the relevant language. This includes everything: syntax, grammar, word choice, pronunciation, etc. Speaking a language with in the most common way is the standard version of that language, nothing else. It is actually possible that there is no regional dialect that speaks that way, but that doesn’t matter. A standard version of a language need not be a regional dialect.<\/p>\n

A standard version of a language is also a necessity if one wants a relatively fonetic spelling system without lots of alternative forms. The idea is that one spells after the sound of the standard version of the language.<\/p>\n

–<\/p>\n

No dialect, however, is more expressive, less corrupt, more logical, more
\ncomplex, or more regular than any other dialect or language. They are sim-
\nply different. More precisely, dialects represent different set of rules or lexical
\nitems represented in the minds of its speakers. Any judgments, therefore, as to
\nthe superiority or inferiority of a particular dialect or language are social judg-
\nments, which have no linguistic or scientific basis.
\nTo illustrate the arbitrariness of \u201cstandard usage,\u201d consider the English r-drop
\nrule discussed earlier. Britain\u2019s prestigious RP accent omits the r in words such
\nas \u201ccar,\u201d \u201cfar,\u201d and \u201cbarn.\u201d Thus an r-less pronunciation is thought to be better
\nthan the less prestigious rural dialects that maintain the r. However, r-drop in the
\nnortheast United States is generally considered substandard, and the more pres-
\ntigious dialects preserve the r, though this was not true in the past when r-drop
\nwas considered more prestigious. This shows that there is nothing inherently bet-
\nter or worse about one pronunciation over another, but simply that one variant is
\nperceived of as better or worse depending on a variety of social factors.<\/p>\n

I don’t care about the typical purist stuff like ‘corruption’, but they are certainly wrong that some dialects are not more complex or regular than others. I really don’t know what makes people make these claims when they are so obviously false. I’ll give a very brief example. Consider a language that has a verb. As it happens, this verb is irregular in one dialect and not so in another. If everything else is equal, then clearly the one dialect is more regular than the other (and less complex), and indeed, better.<\/strong><\/p>\n

Their illustration is strange. First they say that they want to illustrate it, but then end up concluding that their example “shows that there is nothing inherently better or worse about one pronunciation over another, but simply that one variant is perceived of as better or worse depending on a variety of social factors” which is either trivially true becus of the clause about “social factors” (such clauses are almost never explained, in typical sociology fashion), or false becus these differences matter. If the difference is such that other speakers of the language from other dialects fail to understand one, then that is indeed worse, since the purpose of language is generally to be able to communicate. Obviously, if one is not trying to communicate with everyone using the language, this point is irrelevant.<\/p>\n

–<\/p>\n

Constructions with multiple negatives akin to AAE He don\u2019t know nothing are
\ncommonly found in languages of the world, including French, Italian, and the
\nEngl ish of Chaucer, as i l lustrated in the epigraph from The Canterbury Tales. The
\nmultiple negatives of AAE are governed by rules of syntax and are not illogical.<\/p>\n

While perhaps not ‘illogical’, they are redundant and so increase the complexity of a language without adding any increased expressiveness. This is a bad thing.<\/p>\n

–<\/p>\n

The authors spend some time discussing various differences between african american english (AAE) and standard american english (SAE). Some of these differences have relevance to complexity and expressive power, but i’m not knowledgeable enuf to comment on all of their points.<\/p>\n

–<\/p>\n

The first\u2014the whole-word approach\u2014teaches children to recognize a vocab-
\nulary of some fifty to one hundred words by rote learning, often by seeing the
\nwords used repeatedly in a story, for example, Run, Spot, Run from the Dick
\nand Jane series well-known to people who learned to read in the 1950s. Other
\nwords are acquired gradually. This approach does not teach children to \u201csound
\nout\u201d words according to the individual sounds that make up the words. Rather,
\nit treats the written language as though it were a logographic system, such as
\nChinese, in which a single written character corresponds to a whole word or
\nword root. In other words, the whole-word approach fails to take advantage
\nof the fact that English (and the writing systems of most literate societies) is
\nbased on an alphabet, in which the symbols correspond to the individual sounds
\n(roughly phonemes) of the language. This is ironic because alphabetic writing <\/strong>
\nsystems are the easiest to learn and are maximally efficient for transcribing any <\/strong>
\nhuman language. <\/strong>(my bolding)<\/p>\n

So much for their language relativism.<\/p>\n

–<\/p>\n

Chapter 12<\/h3>\n

Another simplification is that the \u201cdead ends\u201d\u2014languages that evolved and
\ndied leaving no offspring\u2014are not included. We have already mentioned Hittite
\nand Tocharian as two such Indo-European languages. The family tree also fails
\nto show several intermediate stages that must have existed in the evolution of
\nmodern languages. Languages do not evolve abruptly, which is why comparisons
\nwith the genealogical trees of biology have limited usefulness. Finally, the dia-
\ngram fails to show some Indo-European languages because of lack of space.<\/p>\n

The authors give the impression that in biology, species do somehow evolve abruptly. But they do no such thing. The analogy works fine in that area. The main problem with the analogy is that languages can share ‘genes’ (words, etc.) between ‘species’ and this does not generally happen in biology. (At least, except for in bacteria<\/a>?)<\/p>\n

–<\/p>\n

\u00a0The term sound writing is sometimes used in place of alphabetic writing, but
\nit does not truly represent the principle involved in the use of alphabets. One-
\nsound \u2194 one-letter is inefficient and unintuitive, because we do not need to
\nrepresent the [p\u02b0] in pit and the [p] in spit by two different letters. It is confusing
\nto represent nonphonemic differences in writing because the sounds are seldom
\nperceptible to speakers. Except for the phonetic alphabets, whose function is
\nto record the sounds of all languages for descriptive purposes, most, if not all,
\nalphabets have been devised on the phonemic principle.<\/p>\n

This is a good observation. I hadn’t thought of that. I shud update my Lyddansk to fix the fonetic principle<\/em> to the fonemic principle<\/em> (in danish ofc). Another way of putting it in ordinary language is: one sound\u2194one symbol, but include only differences in sounds that are relevant.<\/p>\n

–<\/p>\n

If writing represented the spoken language perfectly, spelling reforms would
\nnever have arisen. In chapter 6 we discussed some of the problems in the En\u00a0 glish
\northographic system. These problems prompted George Bernard Shaw to observe
\nthat:<\/p>\n

[I]t was as a reading and writing animal that Man achieved his human
\neminence above those who are called beasts. Well, it is I and my like who
\nhave to do the writing. I have done it professionally for the last sixty
\nyears as well as it can be done with a hopelessly inadequate alphabet
\ndevised centuries before the English language existed to record another
\nand very different language. Even this alphabet is reduced to absurdity
\nby a foolish orthography based on the notion that the business of spelling
\nis to represent the origin and history of a word instead of its sound and
\nmeaning. Thus an intelligent child who is bidden to spell debt, and very
\nproperly spells it d-e-t, is caned for not spelling it with a b because Julius
\nCaesar spelt the Latin word for it with a b.<\/p>\n

The source of the quote is given as: Shaw, G. B. 1948. Preface to R. A. Wilson, The miraculous birth of language.<\/p>\n

Anyway, this particular etymology is actually wrong too! There are many<\/a> such false etymologies that people have based their spelling on. Very utterly foolish. Quoting Wikipedia<\/a>:<\/p>\n

From the 16th century onward, English writers who were scholars of Greek<\/a> and Latin literature<\/a> tried to link English words to their Graeco-Latin counterparts. They did this by adding silent letters to make the real or imagined links more obvious. Thus det<\/em> became debt<\/em> (to link it to Latin debitum<\/em>), dout<\/em> became doubt<\/em> (to link it to Latin dubitare<\/em>), sissors<\/em> became scissors<\/em> and sithe<\/em> became scythe<\/em> (as they were wrongly thought to come from Latin scindere<\/em>), iland<\/em> became island<\/em> (as it was wrongly thought to come from Latin insula<\/em>), ake<\/em> became ache<\/em> (as it was wrongly thought to come from Greek akhos<\/em>), and so forth.[5]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n

–<\/p>\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

Victoria Fromkin, Robert Rodman, Nina Hyams – An Introduction to Language I thought i better read a linguistics textbook before i start studying it formally. Who wud want to look like a noob? ;) I have not read any other textbook on this subject, but i think it was a fairly typical okish textbook. Many […]<\/p>\n","protected":false},"author":17,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[22,1660],"tags":[],"class_list":["post-2793","post","type-post","status-publish","format-standard","hentry","category-language-philosophy","category-linguisticslanguage","entry"],"_links":{"self":[{"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/posts\/2793","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/users\/17"}],"replies":[{"embeddable":true,"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/comments?post=2793"}],"version-history":[{"count":23,"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/posts\/2793\/revisions"}],"predecessor-version":[{"id":2873,"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/posts\/2793\/revisions\/2873"}],"wp:attachment":[{"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/media?parent=2793"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/categories?post=2793"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/emilkirkegaard.dk\/en\/wp-json\/wp\/v2\/tags?post=2793"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}