Complexity in Language.

The Rise and Fall of the English Sentence by Julie Sedivy is one of those long, meaty articles that make too many points to summarize briefly, so I’ll just quote a few bits and urge you to read the whole thing:

Languages with very simple sentence structure are, for the most part, oral languages. It’s the languages that have a culture of writing, developed over a long span of time, that display a fondness for stacking clauses onto one another to create towering sentences. This pattern raises the possibility that the invention of writing, a very recent innovation tagged on to the very last millennia of human evolution, can dramatically alter a language’s linguistic niche, spurring the development of elaborate sentence structure, and leading to the shedding of other features, on a timescale that cannot be achieved through biological evolution. If that’s so, then the languages that many of us have grown up with are very different from the languages that have been spoken throughout the vast majority of human existence.

* * *

The development of intricate sentences in modern European languages has unfolded slowly. These languages now churn out relative clauses with boundless enthusiasm but their common ancestor, Proto-Indo-European, may have lacked the necessary grammatical tools to produce them at all. According to linguist Guy Deutscher, the earliest clay tablets (about 2500 B.C.) of the ancient language Akkadian reveal few embedded clauses. The same is evidently true of the earliest stages of other ancient written languages such as Sumerian, Hittite, or Greek. Although these languages boasted a profusion of grammatical features suitable for expressing subtle nuances of meaning, and included a variety of fancy word-building techniques, they avoided complicated sentence recursion. When they did combine clauses into larger structures, this technique looked less like Russian dolls, with one clause inside another, and more like beads on a necklace, with one clause added next to another, and resembled this sentence from an old Hittite text (14th century B.C.): “I drove in a chariot to Kunnu, and a thunderstorm came, then the Storm-God kept thundering terribly, and I feared, and […], and …”

The invention of writing sparked certain innovations such that by 1800 B.C., Akkadian texts already exhibited complex sentences that rival the prose of Henry James in their complexity.

* * *

All of this suggests that exposure to literary language is essential for the health of complex recursive sentences in English. If certain structures are too rare in speech to be reliably mastered by learners and passed on, then they may fade out within a community of non-readers. Naturally, this raises the question: Could syntactic complexity in literate languages diminish over time, if new technologies (podcasts, video lectures, and audiobooks) tether language more tightly to speech and its inherent limitations?

In fact, heavily recursive sentences like those found in the Declaration of Independence have already been dwindling in written English (as well as in German) for some time. According to texts analyzed by Brock Haussamen, the average sentence length in written English has shrunk since the 17th century from between 40-70 words to a more modest 20, with a significant paring down of the number of subordinate and relative clauses, passive sentences, explicit connectors between clauses, and off-the-beaten-path sentence structures.

* * *

Oral languages may avoid pushing the limits of syntax not just because they are bound to speech, but also because they have other ways to express complex meanings. Linguists take great pains to point out that languages with simple sentences erupt with complexity elsewhere: They typically pack many particles of meaning into a single word. For example, the Mohawk word sahonwanhotónkwahse conveys as much meaning as the English sentence “She opened the door for him again.” In English, you need two clauses (one embedded inside the other) to say “He says she’s leaving,” but in Yup’ik, a language spoken in Alaska, you can use a single word, “Ayagnia.” (Ayagniuq, in contrast, means “He says he himself is leaving”; Ayagtuq means, more simply, “He’s leaving.”)

The templates for creating such complex words can be unruly, making them seem, to the average English speaker, more like cryptographic problems than words. Linguist John McWhorter offers an astounding example from the Siberian tongue Ket, a language in which verbs take pronoun prefixes to mark who is performing an action.

That Ket example is seriously impressive; go ahead, click through and wrap your mind around it.

Comments

  1. Distinction between oral and written languages seems quite arbitrary.

    Kozin invented a term – “oral literary language” to describe the register of 12th century Mongolian in which large parts of the “Secret History of Mongols” are written.

    You see, it is a historic fact that Mongols had no written language in 12th century, but they had a literary language nevertheless – quite complex and fully developed (by the looks of it, with many centuries of history).

    It was used in diplomatic correspondence, for example, which was composed in this oral literary language and sent with envoys who memorized it word by word. It differed quite significantly from colloquial language – in fact, it appears that considerable skill was needed to understand such oral messages (for ordinary people they sounded like riddles).

    So I would say that the difference the author seeks lies in social complexity, not in writing per se which is just a technical tool.

    Language of Skaldic poems comes to mind as western example – yes, the Vikings did have a kind of writing, but it’s use was rather restricted and their literary language was primarily oral, but so complex that needed decipherment even by native speakers.

  2. “It’s the languages that have a culture of writing, developed over a long span of time, that display a fondness for stacking clauses onto one another to create towering sentences.”

    Early Sanskrit, for instance, is an example of an oral language that likes complexity in sentence structure.

  3. Homeric Greek wasn’t very simple either.

  4. There’s too little concrete information in the article for my taste. I wish she’d at least comment on traditional North American Indian oratory or other texts by skilled speakers. Just as in English anyone can write a postcard but not anyone can write a novel, so in oral languages there are more skilled and less skilled speakers. Unfortunately, oral languages are often documented when skilled speakers have become scarce, or when elevated registers have gone out of use, or when the linguist has not stayed with the language long enough to collect and analyze a more complex mode of expression.

    As one mild counterexample, here’s a story in Chicahuaxtla Triqui, with word-by-word analysis.

  5. David Marjanović says:

    I once read there’s a Samoyedic language with such a literary register that makes a point of accumulating rare verb forms.

  6. Ксёнѕ Фаўст says:

    Having read a little about language typology I have an impression that it’s an areal thing: some languages prefer coordination (which appears simpler), while others are more into subordination in sentence building. Also, there are exotic types of relative clauses which may not be recognized initially as such by a researcher accustomed to European languages. Scarcely written languages certainly do possess ‘advanced’ sentence-building features such as dependent moods, clause nominalizations, applicative voices etc.

    Are polysynthetic languages, the likes of Navajo, Ket or Ojibwe, as unlearnable for a European adult as she makes it sound? So you almost have to learn all the individual forms in a paradigm by rote? I’ve always had the impression there should be a good deal of regular agglutination going on there. And the fusional languages of Europe can indeed be very chaotic in their morphology (take German plurals or genitive endings in Polish with their extremely fuzzy ‘rules’).

    a language in which verbs take pronoun prefixes to mark who is performing an action.

    You don’t say 😮

  7. David Marjanović says:

    Make sure you read the part at the end, about scientific writing!

    I have a few details to add to it:

    The decline in sentence complexity may also be due to scientists no longer trying to write in a register appropriate for literature, which may be part of the same phenomenon as no longer wearing suits & ties to conferences. Alternatively or at the same time, it may be an effect of the increased speed of science: some scientific publications of earlier times were clearly designed to be pleasurable to read; there’s no more time, and often no more space, for that, so scientists now strive to get the point across as quickly and clearly as they can manage. (Then as now, individual abilities and sensibilities differ.)

    The rise of noun compounding may have a few additional sources: calquing from German up to the mid-20th century*; native speakers of Romance languages who overestimate what noun compounding can do because they aren’t used to it; native speakers of Chinese who are used to long strings of morphemes that may or may not be classified as nouns and aren’t used to the stricter word classes of English. I suspect that the last two influences have given us the yeast artificial chromosome, patterned after the less questionable bacterial artificial chromosome.

    * Darwin once ran across leglessness or limblessness (I forgot which) in a translation from German and found it very curious; he wrote about it in one of his many letters. Today it seems to be unremarkable in, say, works on the origin of snakes.

    These languages now churn out relative clauses with boundless enthusiasm but their common ancestor, Proto-Indo-European, may have lacked the necessary grammatical tools to produce them at all.

    What about the relative pronoun *h₁jo-?

    In fact, heavily recursive sentences like those found in the Declaration of Independence have already been dwindling in written English (as well as in German) for some time. According to texts analyzed by Brock Haussamen, the average sentence length in written English has shrunk

    There I wonder if the effect has been as strong in German, because short sentences are authoritatively considered good style in English, while in German that never happened. I’ve even read a style guide that says the natural written German sentence instead has one main clause and one subordinate clause; it illustrates its point by a short quote from a saga that exhausts the reader by being all action-action-action-action – it’s all “they seized Thorgrimm, slew him, and…”

  8. David Marjanović says:

    Are polysynthetic languages, the likes of Navajo, Ket or Ojibwe, as unlearnable for a European adult as she makes it sound? So you almost have to learn all the individual forms in a paradigm by rote? I’ve always had the impression there should be a good deal of regular agglutination going on there.

    Depends.

    Navajo is difficult to classify in terms of broad morphological typology: it relies heavily on affixes—mainly prefixes—like agglutinative languages,[47] but these affixes are joined in unpredictable, overlapping ways that make them difficult to segment, a trait of fusional languages.[48] In general, Navajo verbs contain more morphemes than do nouns (on average, 11 for verbs compared to 4–5 for nouns), but noun morphology is less transparent.[49] Navajo is sometimes classified as a fusional language[48][50] and sometimes as agglutinative or even polysynthetic.[19][51]

    So, while there’s plenty of agglutination in Navajo, there’s also a bunch of morphophonology going on, plus unproductive and outright fossilized prefixes that are used on some verbs but not others.

    Ket is like that, only much worse. Many morphemes consist of a single phoneme to begin with. The West Caucasian family has less morphophonology (on the consonants), but the vast majority of morphemes consist of a single phoneme there.

    Ojibwe is nicely agglutinating, except exceptions. Also, it expresses “spider” as “net.make-PEJORATIVE-CONTEMPTIVE”. ^_^

  9. Andreas Dulson (native German speaker) learned Ket after studying it for 20 years.

    So I suppose, any human language, no matter how hard is it, can be learned by another human being.

  10. Of course, there are many types of complexity. At $EMPLOYER’s Journal Club (the first rule of Journal Club …) last week, the following sentence appeared in the article we were studying:

    In responding patients, mutation and neoantigen load were reduced from baseline, and analysis of intratumoral heterogeneity during therapy demonstrated differential clonal evolution within tumors and putative selection against neoantigenic mutations on-therapy.

    The first time I looked at that, I practically choked on it, but by about the tenth time it made sense. Certainly it is not syntactically complex.

  11. Every language, whatever the nature of its structure, is routinely learnt by infants.

  12. marie-lucie says:

    the Siberian tongue Ket, a language in which verbs take pronoun prefixes to mark who is performing an action.

    So do a number of other languages!

    Among them is Chinook, formerly spoken in the Columbia River valley. But one linguist (I don’t remember who) wrote about “The Chinookan structure of French”, citing examples such as Je te le demande ‘I am asking you [for it]’.

    In French the morphemes are written separately, but the usual pronunciation Ch’tel’demand’ would strike a “Martian” linguist as a single word including three pronoun prefixes ch, te, l ‘1st sing, 2nd sing, 3rd sing’. Add an auxiliary and negative morphemes and you get an even more complex structure, as in (spoken) Je n’ t’ l’ai pas d’mandé ‘I did not ask you [for it]’, with six “prefixes” before the verb, three of them pronouns, interspersed with other morphemes. Very Chinookan, and even more complex!

  13. What about the relative pronoun *h₁jo-?
    According to some treatments of PIE I’ve read (can’t remember my sources), there were no relative pronouns in PIE and *h₁jo- was something else (e.g. an anaphoric pronoun).

  14. marie-lucie says:

    If PIE had no relative pronouns, what were the ancestors of Latin qui and others? interrogatives only?

  15. Je n’ t’ l’ai pas d’mandé
    On phonological and adverb-insertion grounds, however, these are three words: je n’ t’ l’ai, pas, and demandé.

  16. m-l: Exactly. Using interrogatives as relatives is one of the features of Core PIE (excluding Anatolian).

    minus273: Really? [ntlɛ] is a French phonological word?

  17. January First-of-May says:

    Very Chinookan, and even more complex!

    While we’re at that, I’m reminded of an actual (supposedly) Chinook example – Lev Uspensky quotes a word inialudam (or similar – иниалюдам in the Russian text) from “the language of the Chinook Indians”, where every phoneme is said to be its own affix, and the actual root is just /d/; it supposedly translates to “I arrived to give this to her” (in Russian, “я прибыл, чтобы отдать ей это”).

    Is it an actual real example of actual Chinook? I don’t know anywhere remotely near enough about of it to be sure, but Uspensky attributes the example to Sapir, who would have probably known better.

    (But yes, French is sometimes similar. I’m often wondering how do native speakers understand it…)

    (And now I’m wondering what’s the Chinook for “She told me you had been to her, and mentioned me to him” [and similar phrases in the rest of that “poem”]. Or the French, for that matter.)

  18. the usual pronunciation Ch’tel’demand’ would strike a “Martian” linguist as a single word

    Coptic is written this way. The relation between Coptic syntax and Hieroglyphic Egyptian is remarkably parallel to French versus Latin. The day this clicked for me, it made Coptic look a lot more transparent.

  19. The relation between Coptic syntax and Hieroglyphic Egyptian is remarkably parallel to French versus Latin. The day this clicked for me, it made Coptic look a lot more transparent.

    Very interesting!

  20. This appears to be a variation on the “Cicero invented recursion” meme by Fred Karlsson, google “Constraints on multiple center-embedding of clauses” for details.

  21. marie-lucie says:

    January: Give me a little time and I will check Boas’ grammar of (Lower) Chinook.

    Sapir did some fieldwork in one of the dialects, and Boas included a couple of pages by Sapir in his grammar, but I only know one short article by Sapir on the language. Sapir’s notes may have been consulted and quoted by later linguists though.

  22. ə de vivre says:

    Really? [ntlɛ] is a French phonological word?

    I think minus273 was separating words by commas, not spaces. The morphophonological words in “je ne te l’ai pas demandé” would be more along the lines: [ʒn̩təlɛ pa d(ə)mãde] or more colloquially [ʃt(ə)lɛ pa dmãde].

  23. marie-lucie says:

    January: Lev Uspensky quotes a word inialudam (or similar – иниалюдам in the Russian text) from “the language of the Chinook Indians”, where every phoneme is said to be its own affix, and the actual root is just /d/; it supposedly translates to “I arrived to give this to her”

    I can’t find this form in Boas’ grammar, but judging from similar examples it probably means only “I will give it to her”, possibly in a related dialect. The writer (or whoever wrote what the writer copied) probably took this word from a longer sentence and used the complete translation for the single word.

    The part meaning ‘give’ is indeed /d/, but the previous /u/ is a prefix which occurs with the majority of verb stems (some of which are very short, others longer), so speakers would probably think of /ud/ as the verb. Final /am/ means ‘future’. I hesitate to identify the initial prefixes, but they must include pronouns for ‘I’ and ‘her’ and perhaps ‘it’ (as in French!).

    p.s. I think that the use of ю in иниалюдам assumes that Chinook has the same /l/ as English, identified by Russian speakers as their own palatalized one. This interpretation is not valid for Chinook any more than for English.

  24. I know two non-Kets who speak good Ket, so it can be done.

  25. I never quite understood what connection the article was trying to make between polysynthesis and lack of syntactic complexity. Rather than compare Ket and German, wouldn’t it make more sense to compare polysynthetic Hungarian or Turkish and their non-written relatives, or isolating written and non-written Sinitic languages?

  26. The part meaning ‘give’ is indeed /d/

    Clearly Indo-European!

  27. p.s. I think that the use of ю in иниалюдам assumes that Chinook has the same /l/ as English, identified by Russian speakers as their own palatalized one. This interpretation is not valid for Chinook any more than for English.

    Difference of palatalized and non-palatalized /l/ is a major feature for Russian speakers, but as far as transliteration from English goes, there is total confusion of which form to use. There is a possibility of a more subtle difference here. I am pretty sure that overwhelming majority of Russians who did not study phonetics are not conscious about the difference between у and ю (when it follows the palatalized consonant). But the first letter stands for /u/ sound and the latter for /y/.

    Not being conscious doesn’t mean not hearing the difference. The choice of ю over у might be decided by it.

    Disclaimer: I am not even an amateur phonetician. Ask your doctor about possible side effects.

  28. marie-lucie says:

    D.O., the problem arises because the Chinook form as quoted above has gone through other people before reaching January: first a linguist studying Chinook (probably Boas or Sapir, both of whom used a phonetic transcription), then at least one other person who wrote in Russian and used the Russian spelling conventions, which when transcribed into English spelling conventions introduced a palatalization which does not exist in Chinook. There may have been a person in between who copied the Chinook form as a single word but kept the translation of the full sentence in which the form was included, thus giving the impression that the form must be more complex than it actually is.

    LH: The part meaning ‘give’ is indeed /d/ – Clearly Indo-European!

    Oh, I did notice that! At least one person would have been overjoyed but did not get as far as Chinook.

    It would look less IE if the Russian transcription offered followed Boas in using t rather than d (for technical reasons).

  29. Trond Engen says:

    It’s Grimm-shifted!

  30. marie-lucie says:

    Y: I never quite understood what connection the article was trying to make between polysynthesis and lack of syntactic complexity.

    From what I understand, “polysynthetic” can be a property of the clause rather than of the whole sentence (which can include relative or subordinate clauses or whatever else can be embedded). Basically, a polysynthetic clause does with morphology what analytical clauses do with syntax, in the way of indicating things like Objects or Adverbials as part of the verb rather than in separate phrases. But there is no reason that a polysynthetic clause cannot use syntactic means in order to join with other clauses, for instance for relativization or subordination, thus creating complex sentences.

    That said, “polysynthesis” does not seem to be universally defined. A few years ago I gave a paper at a conference in which I casually mentioned that the Salishan languages were polysynthetic. It was not a major point, I was only repeating what I thought was received opinion. Two persons familiar with these languages vigorously protested my use of the term. On the other hand, Sarah Thomason who is probably the best-known Salishanist alive does refer to these languages as polysynthetic.

    Rather than compare Ket and German, wouldn’t it make more sense to compare polysynthetic Hungarian or Turkish and their non-written relatives, or isolating written and non-written Sinitic languages?

    Indeed!

  31. Inialudam: it seems that indeed the original claim belongs to Sapir.

  32. What’s the URL you intended?

  33. I found the article really sloppy in its reasoning and assertions at times. For instance, the judgements on what English speakers might say and the contrast with ‘many of the world’s oral languages’ strike me as patronising.

    An English speaker might say: Would you teach me to make bread?

    But a Mohawk speaker would break this down into several short sentences, saying something like this: It will be possible? You will teach me. I will make bread.

    Really? In the oral English I know, it’s more likely that you’d say “Could you teach me how to make bread?” True, this is not three short sentences, but with the ‘how’ in there (‘Could you teach me? How to make bread?’) it’s one step closer to the primitive-looking Mohawk version.

    In English, you might say: He came near boys who were throwing spears at something.

    A Kathlamet approximation would go like this: He came near those boys. They were throwing spears at something then.

    Really? That’s normal spoken English? In ordinary oral English, I could imagine someone saying “He saw these boys, they were throwing spears at something.” Which isn’t far off the primitive-sounding Kathlamet approximation. This whole approach based on comparing written English with word-by-word glosses of non-English languages strikes me as superficial.

    Speakers within a small, tight-knit community have an immense store of shared knowledge. This allows shortcuts in communicating with each other, so that many aspects of meaning can be left implicit. But these shortcuts are not available to speakers of big, sprawling languages who communicate with each other across divides of culture, experience, and specialized knowledge.

    This reminds me of certain Australian English speakers of the less literacy-conscious variety speaking to speakers of English as a foreign or second language. I can understand them perfectly, but I pity poor non-native speakers listening to the colloquialisms, the round-about expressions, the appeals to a shared background, the use of words in non-dictionary meanings, etc. I can’t think of any examples and any made-up example can hardly do justice but (and please excuse the use of orthography to represent ordinary oral pronunciations):

    You gotta think what the other bloke’s tellin’ ya, I might go in and I’ll be thinkin’, ‘Yeah, this is how it oughta be’, but you know, they might be comin’ from somewhere else and if you don’t think about that you’ll suddenly find no one really knows what the other side’s sayin’ and in the end you gotta go back and start from scratch, ’cause no-one knows what the other bloke’s talkin’ about’

    This is actually a pretty literate hypothetical example. There are a lot of people who speak far more enigmatically than this.

    The final part of the article arguing that the splintering into small specialised sub-disciplines using ‘noun pile extravaganzas’ brings our language close to polysynthetic languages spoken by small communities strikes me as pretty vague and sloppy in its reasoning — especially when she invokes headlinese like state hate crime victim numbers.

    The article is food for thought but appears to be heavily driven by the author’s desire to prove her thesis — “syntactic complexity arose among written languages that served vast and diverse communities and we’re now moving away from that as we splinter into smaller subdisciplines, bringing us closer to small pre-literate communities using syntactically simple polysynthetic languages”. This causes her to make vague, broad, and unsubstantiated assertions that ignore or distort the actual facts.

  34. David Eddyshaw says:

    Sedivy seems to be mixing up two really rather different issues:

    (a) are (traditionally) unwritten languages characteristically syntactically simpler than written? to which the answer is, no, they’re not. Counterexamples abound. Impressions to the contrary frankly mostly reflect inadequate description. Hardly surprising; grammatical analysis above individual clause level is hard, even in familiar, well documented written languages.

    (b) do polysynthetic languages frequently show rather flat clause-level syntax? to which the answer is, yes – but surely that is in no way surprising? The complexity is more at word-level and less at clause level, but that’s exactly what polysynthesis is.

    You could claim that that polysynthesis is more characteristic of unwritten than written languages, but given that written languages consititute only a small fraction of all the world’s languages, and that moreover they are disproportionately skewed to just a few language families, it seems a stretch to conclude any sort of cause and effect relationship on the basis of what data we have.

  35. David Marjanović says:

    “Constraints on multiple center-embedding of clauses”

    That’s an interesting paper!

  36. January First-of-May says:

    I think i also found the claim in (a copy of) a Sapir article: it’s page 35 of this PDF edition of Sapir’s Language, an introduction to the study of speech.
    (In the actual book, as seen on archive.org, it appears on page 73.)

    In footnotes, Sapir qualifies the “Chinook” as being “Wishram dialect”, and notes that, technically, the “it” should really be “him” (because Chinook, like French – or indeed Russian – gives grammatical gender even to inanimate objects, and this particular one has the masculine form).
    Presumably the latter footnote was lost in the transmission between Sapir and Uspensky, or the Russian would’ve ended in something like отдать его ей (Russian его can helpfully be either masculine or neuter).

    He insists on the interpretation of the “future” suffix as “arrive for the purpose of”, however. I suspect it has to do with either dialectal variation, the tense in question being a near-future similar to forms with aller in French (or going to in English, for that matter), or misinterpretation involving the previous somewhere on the transmission to Sapir.

    And a surprisingly fitting quote the same Sapir article (PDF p.22, original p.45)…

    “Again, the English l is unknown in Russian, which possesses, on the other hand, two distinct l­-sounds that the normal English speaker would find it difficult exactly to reproduce — a “hollow,” guttural­like l and a “soft,” palatalized l ­sound that is only very approximately rendered, in English terms, as ly.”

  37. marie-lucie says:

    Bathrobe: A Kathlamet approximation would go like this: He came near those boys. They were throwing spears at something then.

    This has to come from Boas’ collection of Kathlamet Chinook tales, of which I have a well-thumbed and studied copy.

    One problem with tales collected before tape-recorders or other means of recording speech in real time is that when someone is speaking slowly in order for someone else to write everything down, the speaker is probably going to simplify their speech, especially in the narrative portions, which will consist or short, apparently simple sentences which the recording person will also repeat to the speaker to make sure the words are right. In such cases, possible subtleties tend to be avoided, and to disappear both from the original language and the translation.

    Although I have struggled with the language of the Kathlamet tales, it has never been my major area of study, so I won’t try to give examples, but the indigenous language I know best is found in another of Boas’ works, “Tsimshian texts – Nass River Texts”, a volume of 70-odd pages with the English text on top and the original with word-for-word translation below. This was my first contact with the Nisqa’a language of the Nass River (British Columbia), before I met any speakers. I found the English text extremely boring! Short sentences, linked by “and then”, with the actions often occurring in the wrong order. It took me several tries to finish the whole thing. After a few months of learning the language with actual speakers (unfortunately not by immersion), I was able to read the texts with some degree of fluency, without constantly looking at the translation. The texts started to come alive! Not only could I correct many of Boas’ mistakes (he had retranslated the texts after collecting them through Chinook Jargon), but I began to understand the flow of events: Boas had used the same simple past tense when he should have backtracked with the pluperfect, for instance. With several years of continuous study of the language I understood and appreciated the tales more and more. Boas had thought that the texts were “only moderately well-told” compared to similar ones he had collected from a speaker with better English. Let’s say that his translation was not even “moderately” good.

    The moral of the story is that hastily transcribed and translated texts, even from people known as good story-tellers, are rarely models of the best that a poorly known language could offer.

  38. January First-of-May says:

    (My previous comment, on the Sapir article, is awaiting moderation.)

    You could claim that that polysynthesis is more characteristic of unwritten than written languages, but given that written languages consititute only a small fraction of all the world’s languages, and that moreover they are disproportionately skewed to just a few language families, it seems a stretch to conclude any sort of cause and effect relationship on the basis of what data we have.

    …Wait, isn’t Sumerian supposed to be pretty close to being polysynthetic?

    (Yes, I totally found that by googling for a list of polysynthetic languages. I did check it out elsewhere, however, because it sounded unlikely.)

  39. David Eddyshaw says:

    isn’t Sumerian supposed to be pretty close to being polysynthetic

    Given that polysynthesis in the hardcore Mark-Bakerish sense is actually not all that common, it’s interesting that the unequivocally Bakeroid-polysynthetic Classical Nahuatl was not only a widespread interethnic language in one of America’s more elaborate cultures but probably the most “written” of American languages after the European invasions.

    Seems to me that the notion that polysynthesis goes with premodernity in anything but a purely coincidental way would need a lot better evidence than is ever likely to be forthcoming in this age of major language death.

    Though I did read somewhere lately an interesting speculation to the effect that “big” languages might, as it were, have got some of their rougher edges knocked off in the process of being serially adopted by absorbed groups that have lost their own original languages – repeated imperfect acquisition.

    Doesn’t seem to have worked for French …

    There also seems to be an element of question-begging: polysynthesis is lost in the face of the demands of communication in complex societies because it’s just too complex as a linguistic strategy, which just shows how polysynthesis must be more complex than elaborate analytical syntax … oh, wait …no, polysynthesis is too simple to express the demands of complex communication … er …

  40. I know I’m a bit of a nitpicker, but I noticed there are problems with the constituent analysis (the square brackets) of both the Declaration of Independence and the Hammurabi sentence (at the time of my writing).

    > The unpredictable aspects of language, the things you just have to know, may be especially slippery for the adult mind—and there are so many more of these in Ket than in English.

    English has a huge lexicon. Lots of stuff you “just have to know”.

    > [ʒn̩təlɛ pa d(ə)mãde] or more colloquially [ʃt(ə)lɛ pa dmãde].

    Intriguing. My own guess would have been [ʃnətlɛ] for the first chunk, is that allowed? Or [ʒnətlɛ]? (Not sure if [ʒn] works as an onset)

  41. ə de vivre says:

    isn’t Sumerian supposed to be pretty close to being polysynthetic

    It’s more agglutinative than polysynthetic. Verbal morphology is the least well understood (and the coolest!) part of the language, but it’s uncontroversial that they’re marked for subject, direct object, dative object. The standard dedication formula that kings had stamped on bricks was:

    Ninŋirsu-ra Enanatum-e mu-na-n-řu-Ø
    god.name-DAT personal.name-ERG ???-3.HUMAN.DAT-3.HUMAN.ERG-build-3.ABS
    “(the king) Enanatum built (it) for (the god) Ninŋirsu

    Sumerian didn’t have full-blown noun incorporation, but it did have a sort-of productive system of semantic incorporation, with the incorporated object often an inalienably possessed body part and the semantic object in an oblique case. Most verbs of perception and cognition are like this:
    en aratta-k-e gig-e igi bi-n-duḫ-Ø
    lord aratta-GEN-ERG wheat-OBL eye APPL-3.HUMAN.ERG-spread-3.ABS
    The lord or Aratta saw the wheat

    Sometimes the compound verb’s meaning is kinda opaque like “šu teŋ” (hand approach) which means to receive. There’s one late loan word where the Akkadian verb “šūḫuzu” (to burn) was reinterpreted as “šu ḫuz” (hand ḫuz). There are also pairs like “tag” (touch) and “šu tag” (hand touch) where the distinction isn’t really clear. The system gets pretty elaborate and the interpretations more contentious.

    Intriguing. My own guess would have been [ʃnətlɛ] for the first chunk, is that allowed? Or [ʒnətlɛ]? (Not sure if [ʒn] works as an onset)

    ʒ only devoices to ʃ before a voiceless consonant. I recall Étienne saying that schwa-insertion in French so-called pronoun groups has a fair amount of free variation. That said, [ʒnətlɛ] sound weird to me, but I’m not a native speaker.

  42. Greg Pandatshang says:

    Whoa, are ᴘᴇᴊᴏʀᴀᴛɪᴠᴇ and ᴄᴏɴᴛᴇᴍᴘᴛɪᴠᴇ really two different morphemes in Ojibwe?

  43. David Marjanović says:

    That said, [ʒnətlɛ] sound weird to me, but I’m not a native speaker.

    I think it puts too much stress on ne, which should be as unstressed as possible.

    Whoa, are ᴘᴇᴊᴏʀᴀᴛɪᴠᴇ and ᴄᴏɴᴛᴇᴍᴘᴛɪᴠᴇ really two different morphemes in Ojibwe?

    That’s what the Wikipedia article says, illustrated with meanings like “bad canoe” for ᴘᴇᴊᴏʀᴀᴛɪᴠᴇ and “some car or something” (presumably “who cares”) for ᴄᴏɴᴛᴇᴍᴘᴛɪᴠᴇ.

  44. marie-lucie says:

    dainichi:

    > [ʒn̩təlɛ pa d(ə)mãde] or more colloquially [ʃt(ə)lɛ pa dmãde].

    Intriguing. My own guess would have been [ʃnətlɛ] for the first chunk, is that allowed? Or [ʒnətlɛ]? (Not sure if [ʒn] works as an onset)

    “Je ne” is always [ʒən], at least at the beginning of an utterance. About [ʒn̩], it might happen after another word, as in “Ben je n’ sais pas” which is not what you would call a prestigious pronunciation even though it does have the “ne”.

    However, in colloquial speech when the “ne” is omitted, (or in cases where there is no negation) you can have an onset with [ʒ] or [ʃ] before another consonant ([ʃ] only before voiceless ones).

  45. marie-lucie says:

    ə de vivre: I recall Étienne saying that schwa-insertion in French so-called pronoun groups has a fair amount of free variation

    True, within the group, and there is personal and regional variation. For instance, je me le demande ‘I wonder about it’ (lit. I ask myself (about it)” can be je m’ le d’mand’ or j’ me l’ demand’ (I tend to use the second one myself). I think that some people say j’ me le d’mand’.

  46. I can’t think of any “big” language that is truly polysynthetic, although to be sure what features are required to make a language polysynthetic are disputed, and it may simply be a prototypical concept, what with polypersonal (polygenderal, whatever) agreement, noun incorporation, or both. In any case, fusional/agglutinative is independent of polysynthetic/not.

    McWhorter talks about the reduction processes that go on in “big” languages. French isn’t the strongest example, certainly. Better cases are Bengali and Cham (now under a lot of pressure from Vietnamese and no longer “big”), both of which are spoken almost exclusively by the descendants of L2 learners. (The same can be said to a lesser degree about American English.) Western Cham in particular is almost scarily regular, with very little characteristic Austronesian morphology left; it reads like an artificial language, except that it has phonemes that a language constructor would be unlikely to choose.

    Update: It is so utterly astounding that the sequences jəm ləd mãd, jməl dəmãd, jmə ləd mãd are all in free variation!

  47. marie-lucie says:

    JC: It is so utterly astounding that the sequences jəm ləd mãd, jməl dəmãd, jmə ləd mãd are all in free variation!

    This means that only the consonants count in order to produce and understand the pronouns. But it does not mean that any person freely uses all three: as I said, I favour the second one in my own speech.

  48. “Really? That’s normal spoken English? In ordinary oral English, I could imagine someone saying “He saw these boys, they were throwing spears at something.” Which isn’t far off the primitive-sounding Kathlamet approximation. This whole approach based on comparing written English with word-by-word glosses of non-English languages strikes me as superficial.”

    Bathrobe, part of my job deals with written statements from the public that are written in spoken English, if you know what I mean, and most long sentences are in fact just a string of clauses mooshed together, with some punctuation sprinkled here and there incorrectly like lipstick on a pig. Once you read them aloud they are perfectly clear.

  49. Greg Pandatshang says:

    Hmmm, popular journalism tells me that we know that Eskimos know a lot about snow because they have 35 or so words for snow. There’s only roughly one conclusion that I can draw from the fact that Ojibwe has >1 suffixes meaning pejorative~contemptuous.

  50. David Marjanović says:

    I think that some people say j’ me le d’mand’.

    Yes, and Paris has moved on to j’ me l’ d’mand’ with a cluster of three consonants.

  51. marie-lucie says:

    I believe you have indeed heard this, but not all of Paris speaks the same!

  52. @ Jim

    That is as I have always suspected.

    I wondered about the claim (from Marianne Mithun) that “34 percent of clauses in conversational American English are embedded clauses. In Mohawk (spoken in Quebec), only 7 percent are. Gunwinggu (an Australian language) has 6 percent and Kathlamet (formerly spoken in Washington state) has only 2 percent.”

    Frankly, I have trouble with the concept of “embedded clauses”. In the spoken language, “When I got up he was having breakfast” includes the embedded clause “when I got up”, but if that is the extent of embedding involved, we are still a far cry from the complexity that Sedivy is claiming for written languages. Until you have multiple embedded clauses, the complexity is not much greater than having coordinate structures. For example:

    When I got up he was having breakfast.
    I got up and he was having breakfast.

    aren’t so different from each other. It’s only when you start embedding within embeddings that claims of syntactic complexity start to have much meaning.

  53. Greg Pandatshang says:

    Also, whoa, Ket polysynthesis is even worsestronger than Navajo?

    Any suggestions for introductory books or articles on Ket? I didn’t find a lot that looked promising through an initial round of Googling.

  54. marie-lucie says:

    Bathrobe: It’s only when you start embedding within embeddings that claims of syntactic complexity start to have much meaning.

    Yes, and as you pile up embeddings people stop following the thread of your thought. It works best in writing, as the reader can go over complex sentences again and again.

  55. Complex syntactic constructions piled one upon another are one of the most popular genres of folk oral tradition in many cultures.

    “This is the House that Jack built” and all that.

    Mongolian example (rather short, so it can be posted in entirety):

    Manai khaany myangan langiin myangan möngön bömbgiig manai khaany myangan langiin myangan möngön bömbög bish geed öör khenii khaany myangan langiin myangan möngön bömbgiig myangan langiin myangan möngön bömbög gekh yum be?

    If our king’s one thousand silver balls worth one thousand taels are not called our king’s one thousand silver balls worth one thousand taels then whose king’s one thousand silver balls worth one thousand taels can be called one thousand silver balls worth one thousand taels?

  56. January First-of-May says:

    Final embedding, like in House that Jack Built, isn’t actually that complicated. I vaguely recall someone giving an example from, IIRC, the 1001 Nights that went six layers in.

    It’s the other kinds of embeddings that are tricky.

  57. David Eddyshaw says:

    Any suggestions for introductory books or articles on Ket?

    “Ket”, by Edward Vajda, in the LINCOM Languages of the Word/Materials series.

  58. David Eddyshaw says:

    Complex syntactic constructions piled one upon another are one of the most popular genres of folk oral tradition in many cultures.

    In a very similar way, polysynthetic languages can use elaborate polysynthetic constructions specifically as an artistic choice – a deliberate display of linguistic virtuosity. I’ve seen this explicitly stated at least for Nahuatl and Arapaho.

    It may be that the dichotomy (so far as it’s real at all) is not between oral/written but between workaday and elevated diction; not the same thing at all, however much we literate modern people might think so, with our inability to memorise even the simplest holy book, epic poem or comprehensive morphophonemic treatise.

  59. David Eddyshaw says:

    Ainu seems to be a good example of this, come to think of it: complex polysynthesis is very evident in the oral epics, the yukar, but very much less so in ordinary spoken language (though I suppose you have to allow also for the effects of language obsolescence there, which confuses the issue.)

  60. It may be that the dichotomy (so far as it’s real at all) is not between oral/written but between workaday and elevated diction; not the same thing at all, however much we literate modern people might think so

    An excellent point.

  61. It’s the other kinds of embeddings that are tricky.

    Centre-embedding is so difficult to process that even in written language third-degree centre-embedding seems to be the greatest complexity ever observed in the wild (according to Karlsson 2007). In spoken language, even second-degree centre-embedding is virtually unknown:

    The rat that the cat that the dog chased killed ate the malt.

    [The rat [that the cat [that the dog chased] killed] ate the malt]

    However, as Hurford (2012) points out, educated users of some languages, immersed in a rhetorical tradition which cherishes syntactic complexity (he mentions literary German, in particular), experience more exposure to centre-embedding, and therefore may cope more successfully with such sentences. Humans can reportedly be trained to process up to five levels of centre-embedding in laboratory conditions 😉

  62. David Marjanović says:

    The rat that the cat that the dog chased killed ate the malt.

    I can deal with that if the intonation is just right. But that’s probably the limit.

    a rhetorical tradition which cherishes syntactic complexity (he mentions literary German, in particular)

    German of any kind doesn’t have that much of a rhetorical tradition, though. My impression is it has less than English. Immersion in that tradition is even less common – Americans hold formal debates in highschool classes; we don’t.

  63. [The rat [that the cat [that the dog chased] killed] ate the malt]

    Center embedding corresponds to general recursion in programming languages, whereas final embedding corresponds to tail recursion, which is just iteration by another name. When we try to understand sentences like the above, it’s the string of finite verbs “chased killed ate” that doesn’t work: we try to recall their subjects correctly but fail. In the “House that Jack Built” case, all the ]s are at the end, and the brain has no trouble processing them.

    I thought someone posted a hilarious German center-embedded translation of Jack (by the way, is this the Jack of the Jack tales?), but I can’t seem to find it.

  64. Here is one of Hurford’s German examples (in his original spelling, with no scharfes S):

    Ich weiss, dass der Keiser den Eltern den Kindern Fussball spielen lehren helfer soll.

    According to p.c. from one of Hurford’s German students, it “is not particularly difficult to interpret”. Hurford speculates that the Germans’ ability to cope with centre-embedding is “probably facilitated by the verb-final structure of German subordinate clauses” (in other words, German-speakers are accustomed to waiting patiently for the verb — it will come one day).

  65. David Marjanović says:

    There are enough mistakes in this example that I’m unsure of the interpretation! If it is

    Ich weiß, dass der Kaiser den Eltern die Kinder Fussball spielen lehren helfen soll

    , then it’s straightforward enough that it might well occur spontaneously in the wild on rare occasions, and means “I know that the emperor [?!?] is supposed to help the parents teach the children to play football”.

    German-speakers are accustomed to waiting patiently for the verb

    Bingo.

  66. Mark Twain quotes typical German sentence ending in a verb:

    “But when he, upon the street, the (in-satin-and-silk-covered-now-very-unconstrained-after-the-newest-fashioned-dressed) government counselor’s wife met”

  67. “And that sentence is constructed upon the most approved German model. You observe how far that verb is from the reader’s base of operations; well, in a German newspaper they put their verb away over on the next page; and I have heard that sometimes after stringing along the exciting preliminaries and parentheses for a column or two, they get in a hurry and have to go to press without getting to the verb at all. Of course, then, the reader is left in a very exhausted and ignorant state.”

  68. marie-lucie says:

    SFR: “But when he, upon the street, the (in-satin-and-silk-covered-now-very-unconstrained-after-the-newest-fashioned-dressed) government counsellor’s wife met” (Mark Twain)

    Reading the American press, it seems to me that English syntax is starting to emulate German syntax in the elaboration of the noun phrase in between the article and the noun itself. I have not collected examples, but I am often struck by a new complexity.

  69. > Final embedding, like in House that Jack Built, isn’t actually that complicated. […] It’s the other kinds of embeddings that are tricky.

    Isn’t initial embedding even easier in some cases? I can understand why “The House that Jack Built” is easy to parse, since when the whole sentence ends, necessarily all the nested clauses have to end too. But meaning-wise, is it really easy? As far as I can see, all the relative clauses are restrictive. So first we have “This is the rat… “. What rat? “… the rat that ate the malt … ” What malt? Etc, until we get all the way to the end. Only when we get to Jack, whom we know, do we know what house, and what malt, and what rat we’re talking about.

    That’s not cognitively easy. Because the old info is all the way at the end. Embedding at the beginning with pre-positioned relative clauses (like in e.g. Japanese) would be much easier to understand, since you can start with the old information and move towards the new information. With that kind of embedding the understanding would be as easy as:

    You know Jack? Yeah.
    He built a house. OK.
    Some malt lay there. OK.
    Some rat ate that malt. OK.
    This is that rat! Aha!

    (I used a shortened version of the rhyme for the purpose)

  70. There are enough mistakes…

    Oops, the typos are mine, not Hurford’s. It was past 1 a.m.

  71. spielen lehren helfen soll — I am far from being a native speaker of German, but I have no problems parsing this infinitive pileup. Your syntax tree may be center-embedded, but it feels more like I am processing it as a single compound verb form with four NP arguments. Is that what a serial verb does?

  72. spielen lehren helfen soll

    Nothing formally wrong with that, but only an intellectual putting on the Ritz would actually say or write that nowadays.

    Decades ago I learned by example to say soll spielen lehren helfen, and I still say it for considered reasons. Most people don’t do that, though, but prefer to avoid a pile-up altogether.

    The woman in the street thus will say something like soll helfen, sie spielen zu lehren (“sie” has been pulled inwards to help avoid the pile-up) or, more colloquially, soll helfen, ihnen [das] Spielen beizubringen (note that the verb “spielen” has mutated into the noun “Spielen”).

    “Sie lehren” is a bit unusual in the Germany-German of my neck of the woods, which is why I replaced it with “ihnen beibringen”. I seem to recall that David M’s Austria-German uses “lehren” more often for “teach”.

  73. Another Mark Twain quote:

    “An average sentence, in a German newspaper, is a sublime and impressive curiosity; it occupies a quarter of a column; it contains all the ten parts of speech — not in regular order, but mixed; it is built mainly of compound words constructed by the writer on the spot, and not to be found in any dictionary — six or seven words compacted into one, without joint or seam — that is, without hyphens; it treats of fourteen or fifteen different subjects, each inclosed in a parenthesis of its own, with here and there extra parentheses which reinclose three or four of the minor parentheses, making pens within pens: finally, all the parentheses and reparentheses are massed together between a couple of king-parentheses, one of which is placed in the first line of the majestic sentence and the other in the middle of the last line of it — after which comes the VERB, and you find out for the first time what the man has been talking about; and after the verb — merely by way of ornament, as far as I can make out — the writer shovels in “haben sind gewesen gehabt haben geworden sein,” or words to that effect, and the monument is finished. “

  74. David Marjanović says:

    Your syntax tree may be center-embedded, but it feels more like I am processing it as a single compound verb form with four NP arguments.

    Good idea. Maybe that’s what’s going on in my head.

    Decades ago I learned by example to say soll spielen lehren helfen

    That’s widespread, IIRC even prescribed by a bunch of prescriptivists. But it’s not geographically universal; it feels downright ungrammatical to me (in Standard and dialect).

    But yes, usually such pileups are avoided exactly as you describe.

    I seem to recall that David M’s Austria-German uses “lehren” more often for “teach”.

    Not at all! It’s a purely literary word. Beibringen is usual, and colloquially you can also encounter lernen (I hear prescriptivists howling and wailing in the distance); then there’s unterrichten for teaching in a formal setting without having to mention who the learners are.

    An average sentence, in a German newspaper,

    Such extremes did occur in scientific writing in Twain’s time; and the newspapers (or some of them?) were aspiring to literary merits, so maybe you’d find half that in a newspaper from back then…

  75. the writer shovels in “haben sind gewesen gehabt haben geworden sein,” or words to that effect,

    “Or words to that effect” cracks me up.

  76. As Stu notes, these very long, subclause-ridden sentences Twain satirized (what I call the Thomas-Mann-style of writing) have gone out of fashion. Legalese and admistrative writing still have complicated nested constructions, as part of their overall endeavour to be penetrable only to the specislists, but that’s the same all over the world.

  77. C’mon, Hans, Uncle Thomas never wrote like that. It is merely the case that, in order to get something out of his novels, one must be high up the pole of German-across-the-centuries. He was a consummate show-off, but that’s another story.

  78. @Stu – How about these sentences? In my book, they’re very long and sub-clause ridden, and I could easily find more like them:
    “Die Liebe, die Schridaman zu Sumantras goldfarbenem Kinde hatte, nachdem er sie im frommen Bade belauscht, – eine Liebe, so feurig-ernst, dass sie für ihn, zu Nandas populärer Erheiterung, die Gestalt einer Krankheit zum Tode und der Überzeugung, sterben zu müsse, angenommen hatte, – diese heftige, leidende und im Grunde zartsinnige Ergriffenheit also, entzündet durch ein reizendes Bild, dem er jedoch sogleich die Würde der Perso zu wahren bestrebt gewesen war, – kurzum, diese aus der Vermählung von Sinnenschönheit und Geist geborene Begeisterung war, wie sich versteht, eine Sache seiner gespannten Selbstheit, – vor allem aber doch eine Sache seines brahmanischen, von der Göttin ‘Rede’ mit Gedankeinbrunst und Einbildungskraft begabten Hauptes gewesen, welchem der ihm anhängende milde Körper, wie das in der Ehe deutlich geworden sein mochte, keine ganz ebenbürtige Gesellschaft dabei geleistet hatte.”
    Or this:
    “Als sie nun einen Tag und einen halben gereist waren, unter Menschen auf Landstraßen wie auch allein durch Wälder und Einöden, wobei jeder seine Wegeslast auf dem Rücken trug: Nanda einen Kasten mit Betelnüssen, Kaurimuscheln und auf Baspapier aufgetragenem Altarot zum Schminken der Fußsohlen, womit er das Roherz der Tiefstehenden zu bezahlen gedachte, und Schridaman die in ein Rehfell eingenähten Gespinste, die aber Nanda aus Freundschaft auch zuweilen noch aufhuckte, kamen sie an einen heilgen Badeplatz Kâlîs; der Allumfangenden, der Mutter aller Welten und Wesen, die Vischnus Traumtrunkenheit ist, am Flüsschen ‘Goldfliege’, das fröhlich wie eine losgelassene Stute aus der Berge Schoß kommt, dann aber seinen Lauf mäßigt und an heiliger Stelle sanft mit dem Strome Djamna zusammenfließt, der seinerseits an überheiliger Stelle in diewige Ganga mündet, – dieser aber mündet viefach ins Meer.”
    (From “Die vertauschten Köpfe”)

  79. Gott hilf uns! I got a few clauses into the first sentence before giving up in despair, and I know now that I will never attempt to read Mann in the original.

  80. Hans, Die vertauschten Köpfe is deliberately high-flown. I read it for the first time only last year. Compare and contrast Lotte in Weimar, or Zauberberg.

    Steve, it must be because you balk too often at syntax and words you’re not completely at home with. When you just read along – and have a suitably raised consciousness, as when reading any high-falutin’ prose – all is well. It’s supreme artifice, man ! Bureaucrats only manage artificial.

  81. Oh, well, the Anglosphere has Ruskin:

    And round the walls of the porches there are set pillars of variegated stones, jasper and porphyry, and deep-green serpentine spotted with flakes of snow, and marbles, that half refuse and half yield to the sunshine, Cleopatra-like, ‘their bluest veins to kiss’ — the shadow, as it steals back from them, revealing line after line of azure undulation, as a receding tide leaves the waved sand; their capitals rich with interwoven tracery, rooted knots of herbage, and drifting leaves of acanthus and vine, and mystical signs, all beginning and ending in the Cross; and above them, in the broad archibolts, a continuous chain of language and of life — angels, and the signs of heaven, and the labours of men, each in its appointed season upon the earth; and above these, another range of glittering pinnacles, mixed with white arches edged with scarlet flowers, — a confusion of delight, amidst which the breasts of the Greek horses are seen blazing in their breadth of golden strength, and the St. Mark’s lion, lifted on a blue field covered with stars, until at last, as if in ecstasy, the crests of the arches break into a marble foam, and toss themselves far into the blue sky in flashes and wreaths of sculptured spray, as if the breakers on the Lido shore had been frost-bound before they fell, and the sea-nymphs had inlaid them with coral and amethyst.

  82. Count Leo won’t be outdone:

    В то время, когда на юбилее московского актера упроченное тостом явилось общественное мнение, начавшее карать всех преступников; когда грозные комиссии из Петербурга поскакали на юг ловить, обличать и казнить комиссариатских злодеев; когда во всех городах задавали с речами обеды севастопольским героям и им же, с оторванными руками и ногами, подавали трынки, встречая их на мостах и дорогах; в то время, когда ораторские таланты так быстро развились в народе, что один целовальник везде и при всяком случае писал и печатал и наизусть сказывал на обедах речи, столь сильные, что блюстители порядка должны были вообще принять укротительные меры против красноречия целовальника; когда в самом аглицком клубе отвели особую комнату для обсуждения общественных дел; когда появились журналы под самыми разнообразными знаменами, – журналы, развивающие европейские начала на европейской почве, но с русским миросозерцанием, и журналы, исключительно на русской почве, развивающие русские начала, однако с европейским миросозерцанием; когда появилось вдруг столько журналов, что, казалось, все названия были исчерпаны: и “Вестник”, и “Слово”, и “Беседа”, и “Наблюдатель”, и “Звезда”, и “Орел”, и много других, и, несмотря на то, все являлись еще новые и новые названия; в то время, когда появились плеяды писателей, мыслителей, доказывавших, что наука бывает народна и не бывает народна и бывает ненародная и т. д., и плеяды писателей, художников, описывающих рощу и восход солнца, и грозу, и любовь русской девицы, и лень одного чиновника, и дурное поведение многих чиновников; в то время, когда со всех сторон появились вопросы (как называли в пятьдесят шестом году все те стечения обстоятельств, в которых никто не мог добиться толку), явились вопросы кадетских корпусов, университетов, цензуры, изустного судопроизводства, финансовый, банковый, полицейский, эманципационный и много других; все старались отыскивать еще новые вопросы, все пытались разрешать их; писали, читали, говорили проекты, всё хотели исправить, уничтожить, переменить, и все россияне, как один человек, находились в неописанном восторге.

  83. Very good, Piotr. I had forgotten about Ruskin. Now Carlyle springs to mind – not so much for nested clauses as for generally whacked-out prose.

  84. Nabokov is a contender, too; here are three consecutive sentences from Дар [The Gift]:

    Опытным взглядом он искал в ней того, что грозило бы стать ежедневной зацепкой, ежедневной пыткой для чувств, но, кажется, ничего такого не намечалось, а рассеянный свет весеннего серого дня был не только вне подозрения, но еще обещал умягчить иную мелочь, которая в яркую погоду не преминула бы объявиться; все могло быть этой мелочью: цвет дома, например, сразу отзывающийся во рту неприятным овсяным вкусом, а то и халвой; деталь архитектуры, всякий раз экспансивно бросающаяся в глаза; раздражительное притворство кариатиды, приживалки, – а не подпоры, – которую и меньшее бремя обратило бы тут же в штукатурный прах; или, на стволе дерева, под ржавой кнопкой, бесцельно и навсегда уцелевший уголок отслужившего, но не до конца содранного рукописного объявленьица – о расплыве синеватой собаки; или вещь в окне, или запах, отказавшийся в последнюю секунду сообщить воспоминание, о котором был готов, казалось, завопить, да так на углу и оставшийся – самой за себя заскочившею тайной. Нет, ничего такого не было (еще не было), но хорошо бы, подумал он, как-нибудь на досуге изучить порядок чередования трех-четырех сортов лавок и проверить правильность догадки, что в этом порядке есть свой композиционный закон, так что найдя наиболее частое сочетание, можно вывести средний ритм для улиц данного города, – скажем: табачная, аптекарская, зеленная. На Танненбергской эти три были разобщены, находясь на разных углах, но может быть роение ритма тут еще не настало, и в будущем, повинуясь контрапункту, они постепенно (по мере прогорания или переезда владельцев) начнут сходиться: зеленная с оглядкой перейдет улицу, чтобы стать через семь, а там через три, от аптекарской – вроде того, как в рекламной фильме находят свои места смешанные буквы, – причем одна из них напоследок как-то еще переворачивается, поспешно встав на ноги (комический персонаж, непременный Яшка Мешок в строю новобранцев); так и они будут выжидать, когда освободится смежное место, а потом обе наискосок мигнут табачной – сигай сюда, мол; и вот уже все стали в ряд, образуя типическую строку.

  85. Hans, Die vertauschten Köpfe is deliberately high-flown. I read it for the first time only last year. Compare and contrast Lotte in Weimar, or Zauberberg.
    Well, you said he never wrote like that, and while he was aiming for a Brahmin kind of elaboration in his prose in “Köpfe”, I could find such sentences in his work elsewhere, if I felt like going through it right now and if I wouldn’t have blisters on my fingers already from typing the examples I gave. 🙂
    And sure, writers in other languages can be as long-sentenced, as has been shown.

  86. The Polish writer Jerzy Andrzejewski wrote a short novel entitled Bramy raju (The Gates of Paradise, 1960), which consists of two sentences. The first is about forty thousand words long, and the second is, “And they marched all night.” Of course Andrzejewski cheated a little, since many of his commas, dashes and semicolons could be replaced with full stops. For similar reasons, various forms of stream of consciousness (like Molly’s soliloquy in Ulysses) should be disqualified. They are not extraordinarily complex — just amorphous and underpunctuated.

  87. per incuriam says:

    I know now that I will never attempt to read Mann in the original

    Buddenbrooks and Zauberberg are both eminently readable.

  88. Buddenbrooks and Zauberberg are both eminently readable.

    As is Felix Krull (if you don’t mind reading an unfinished novel); I don’t want to dissuade anyone from reading Mann!

  89. I read the novella Mario und der Zauberer at university, possibly because it was felt to be a suitable introduction to Mann for students. I remember it as having a heavy style compared with other things we read (e.g., Eichendorff and C. F. Meyer) but it wasn’t impossible, even for a first year student. It also had an interesting message.

  90. As (I’m told) the German version of an Albee play has it, “Wer hat Angst vor Thomas Mann?”

  91. Henry James (who is, in fact, quite readable still) meets some Python scripts.

  92. marie-lucie says:

    I have not tried to read much Mann, but I found Kafka very readable, even with my limited instruction in the language.

  93. Me too.

  94. Henry James talking in Jamesian (according to Edith Wharton).

  95. Sedivy.

    Would this name be pronounced as the ‘word’ in the line ‘a kiddley divey too, wooden shoe’?

    Or is the stress on the first syllable?

  96. Good question. It’s a Czech name (from šedivý ‘grey’), so my guess would be first-syllable stress, but who knows how it’s been naturalized?

Trackbacks

  1. […] View original: Complexity in Language. […]

Speak Your Mind

*