1097: Assimilation (s.s.w.2) Dec 10, 2017

It might not seem that assimilation—when one sound becomes similar to or completely elides with one that is adjacent to it—would have too much to do with syllables, but there are similarities. Rules for assimilation, in some language, are based upon sonority hierarchy, such as in Finnish in which a less sonorous phoneme can elide with a more sonorous one it precedes. For instance, [tn] could become [n]. Plenty of other languages use assimilation, including the English 'sandwich' which can be pronounced idyllically as [ˈsændˌwɪt͡ʃ], but may more likely appear as /ˈsænˌwɪt͡ʃ/ or /ˈsæmˌwɪt͡ʃ/ (with only [n] or alternatively [m]). In languages like Finnish however, because it follows the same patterns as the sonority hierarchy, this would change how many syllables are in a given word. Nevertheless, English has historically had assimilation (which is called diachronic assimilation) that changed the quality and number of syllables in words that are now consistently a certain amount of syllables. For instance, the '-ea-' used in orthography now often represents the way that people used to speak, such as in 'sea' which would not have rhymes with 'see' and would have had two syllables.

1096: Defining Syllables: the MOP and SSP (s.s.w.1) Dec 9, 2017

This post marks the 3rd year of Word Facts', and the start of Syllable and Stress Week (SSW). Thank you for the support

There are two principles that allow people to determine how a word is broken up into syllables, at least generally in English. First, there is the Sonority Sequencing Principle (SSP) which was discussed here last week, and then there is the Maximum Onset Principle (MOP). Simply put, the MOP states that, between vowels, consonants are assigned to be the onset i.e. it is the first part of a syllable along with the vowel it precedes. It is for this reason that it has been found that people will almost certainly divide tatatata... as ta•ta•ta•ta... as opposed to tat•at•at•at.... Nevertheless, there are language-specific constraints on this, which is why syllables can end in consonants, because, according to the SSP, if a phoneme's sonority (how loud it is) falls between what is immediately lower in sonority (possibly the onset: optional) and what is higher in sonority (possibly the nucleus, usually a vowel), or vice versa (meaning the final phoneme is possible a coda: optional) that phoneme will not be part of a new syllable. If that is confusing, there is a graphic from bluelook.net that should hopefully clear up what those peaks and valleys look like in syllables, but also this will be explained further in the next few days.

1095: Dropping Fricatives if e'er They're in the Middle Dec 8, 2017

People who have studied literature are probably familiar with the form of 'e'er' for 'ever', or other examples—particularly in poetry—where the fricative (usually [v]) is dropped from the middle of a word. This can be seen repeatedly in the first four lines of An EPITAPH by Joseph Giles(*):
If e'er sharp sorrow from thine eyes did flow,
If e'er thy bosom felt another's woe,
If e'er fair beauty's charms thy heart did prove,
If e'er the offspring of thy virtuous love...
By the eighteenth century, this was fairly common in poetry, though not as popular in everyday speech. What was becoming more common at that time was to drop a fricative from the contracted forms, so 'isn't' might become 'in't', which is still used by some people as an alternative form of 'ain't, but also this occurred with 'wasn't and 'hasn't' and many other words. As is always the case with the evolution of language, it was faced with some protest, but there are still some traces today.
Tomorrow is the 3-year anniversary of Word Facts, and the start of a week of posts covering syllables and stresses.

(*) A Collection of Poems in Six Volumes. By Several Hands. Vol. V. London: printed by J. Hughs, for R. and J. Dodsley, 1763 [1st ed. 1758], pp. 92-93. 6v.: music; 8⁰. (ESTC T131163; OTA K104099.005). Retrieved from http://www.eighteenthcenturypoetry.org

1094: Isopsephy Dec 7, 2017

Several cultures have used letters to represent numerals. Famously, the Romans and other Europeans used Romans numerals, which assigns numbers to letters, but quite often, people would do it the other way around. Isopsephy is a Greek word denoting the practice of taking the sum of the numbers assigned to a word, but this requires that every letter in a given writing-system has a numeric value. This practice was Greek, but other cultures with other writing systems did this as well, including with the Roman alphabet, the Hebrew alephbet, and the Arabic abjad. Some people used this for prophesies, while other people would use this alphanumeric system as a code of sorts. While certainly less popular today, the notion that the number of the beast is 666 or, historically, 616 comes from this practice, but given that different people assigned different numbers to letters in both the same and different writing-systems, the same word will often yield varying results.

1093: Negative Forms with Distinct Pronunciations Dec 6, 2017

For many people—particularly outside of the U.K.—the vowel in the negative contraction of 'can', 'can't', is the same one, either [a] or [æ], as is used for the positive form. This should make sense, since with the exception of 'do', none of the commonly used words with a negative contractional form change their vowels, but this was not always the case. In Middle English, many verbs that were irregular would have a phonetically distinct negative form. Then, 'can' was pronounced [kæn] like its pronounced in Standard American English today, but 'can't was pronounced as [kɑ:nt]. For this word, there is a similar situation that happens in Received Pronunciation, or other souther English dialects, but this occurred in many more words at the time. Not only was the vowel different, but one the positive form was short while the negative form was long.

1092: Trilling: ɾ (pot of tea and para ti) Dec 5, 2017

Standard American English has the sound [ɾ], called a tap or a flap, in words like 'city' [sɪɾi] which in Received Pronunciation would be with a [t] ([sɪti]) or would be a glottal stop in some other dialects in Britain [sɪʔi]. This [ɾ] sound is produced by placing one's tongue on the roof of the mouth (specifically the alveolar ridge) for less time than one would to pronounce [t] as in 'tango'. This process is also sometimes called trilling. Trilling does not exist just as an allophone of [t], but also is an allophone for [r] or [ɹ], or a whole bunch of other sounds, theoretically. Indeed, in Spanish, the 'rolled R' or 'trilled R' written orthographically with a double-R in the middle of words is the same phoneme as this. In fact, the phrase "pot of tea" in Standard American English is more or less phonetically identical to the Spanish "para ti", which would be transcribed [paɾati]. The reason that IPA [ɾ] looks more like a lower-case R and less like a T is because of this in Spanish.

1091: Allophones (Articulation) Dec 4, 2017

As discussed yesterday, if a phoneme consistently morphs into a different sound in a consistent phonetic context, this means that the sound which it becomes is an 'allophone'. Another example from English that was not looked at yesterday was that the glottal stop [ʔ] and the tap [ɾ] are considered allophones of [t] because they appear as the onset for certain syllables (which will be discussed specifically during Word Facts' Syllable and Stress Week starting December 9th) such as in 'butter': [bʊɾɚ] in Standard American English and [bʊʔə] in some northern dialects of British English. There all should make sense as allophones, as they all (including the examples from yesterday) share the same manner of articulation (row on the chart below) or place of articulation (column on the chart below).

It may sound odd at first, but in kiSwahili, [d] as in 'delta' is considered an allophone of [l] as in 'lima'. This is because, before an [n], [l] becomes a [d], such as in 'ulimi' meaning 'language' or 'message', which, when pluralized with 'n-' becomes 'ndimi'. Not only that, but also [l] and [d] share the same manner of articulation.

1090: Allophones (Definition) Dec 2, 2017

There are certain pairs of phonemes (sounds) that—in very specific contexts—would probably be indistinguishable in English, such as [p] or [b] in 'spot' (or 'sbot'). Nevertheless, the phoneme [p] can represent the sound produced by 'p' in 'pot', 'spot', and 'stop', even though they are all slightly different. While [b] and [p] may not be distinguishable in a few words, there are occasions when the difference changes the meaning, such as 'bin' and 'pin'. However, none of those different words before that have a sound represented by 'p' will change or lose meaning (in English) if one is used instead of the other. These are called 'allophones', because they all act as the same sound, in that one can be substituted for the other to an odd-sounding effect but identical meaning. In other words, these allophones are not part of a language's phoneme inventory. 

Another criteria for allophones, however, is that if one sound consistently assimilates to another in some given context, it is an allophone. Therefore, even though the sound [t͡ʃ] (like the CH in 'church') can create different meaning when it replaces [t] such as in 'porch' and 'port', because a T [t] before an R [ɹ] becomes a [t͡ʃ] (consider 'tap' [tap] and 'trap' [t͡ʃɹap]) [t͡ʃ] is considered an allophone of [t]. It should be noted that allophones change from language to language. There will be more on this tomorrow, considering other languages.

1089: Unstable Vowels Dec 2, 2017

'Do', 'does', and 'don't' are obviously related to each other, as the latter two forms are only versions of the first. Nevertheless, each one of them has a different vowel; in Standard American English, the vowel is 'do' is [u], in 'does' it is [ʊ], and in 'don't' it is [oʊ̯]. You can see these on the chart (which is modelled after where a tongue has to move to pronounce the vowel) from speechmodification.com below. These vowels are said to be unstable, and 'do' is not the only word to have this unstable vowel historically, but it is a good example of how these tend to lengthen and lower (see chart and observe that [ʊ] is produced physically lower than [u]). Some of the reason for this is that it can be easier to produce, but in some languages, vowels will change regularly depending upon which consonants potentially follow them; this concept will be explored in the future soon.

1088: Untranslatable Colors Dec 1, 2017

Linguistics requires a lot of scientific analysis, but since there are certain cultural elements that influence the way people speak, unless one who is studying a language understands this, linguists can misunderstand the way that people talk, due to their own biases. For example, some languages have fewer words for colors than others, which is fairly easy to describe, but sometimes comparing words for colors is not so simple. Hanunó’o, a Filipino language, has terms for ‘light’ and ‘dark’ which is understandable from an English-speaking perspective, but colors can also be described as ‘wet’ and ‘dry’, which has no equivalent in English. Without knowing this, it could seem that Hanunó’o has fewer words for colors than it actually does.

1087: S-Selection (Grammatical but Nonsensical) Nov 30, 2017

Neither sentence: "I see he" nor "I run him" can be considered acceptable, but the reason for each is different. In the first one, the problem is that 'he' is only used as a subject, even when it appears in the middle or end of a sentence, but given that 'see' takes an object, it would require 'him' to be used. This issue does not have to relate to what each word means semantically, but how it all fits together syntactically (or not in this case). On the other hand, with "I run him", while it is possible for 'run' to take an object, such as in "I run a race", this sentence does not make sense considering the meaning of each word together. Chomsky's famous sentence—whether you agree with his conclusions—"colorless green ideas sleep furiously" is an example of how people can see that something is grammatical but nonsensical.

1086: Sonority and Syllables Nov 29, 2017

Sounds are not created the same way physically, and this not only—though most obviously—changes the way the sound (phoneme) sounds, but also it changes how loud it will be on average. Vowels, for instance, are the loudest while stops (such as [t] or [b] are the least sonorous. All phonemes can be placed on the sonority hierarchy which goes, from loudest to quietest: vowels, approximates (like [l] or [r]), nasals (like ([m] and [n]), fricatives (like [f]), affricates (like [d͡ʒ], or the J in 'jump') and the last category is stops. Aside from simply being an amusing fact that some sounds are generally louder than others, it is because of this hierarchy that syllables exist, because when one speaks, as the sounds decrease and then increase in volume, more or less accidentally or at least coincidentally, this creates—simply put—the rhythm of syllables. This is a rich topic, and this hardly at all scratches the surface, so expect more in the future soon.

1085: Determiners are not Adjectives Nov 28, 2017

For a long period of time, articles ('the', 'a', or 'an') were considered to be a type of adjective rather than a type of determiner. Though they do act differently in some ways, they do—at least at first glance—seem to modify nouns and though they have to go before adjectives, there is already an order to adjectives anyway. The difference is not only in the way it operates syntactically, but also differences semantically. Determiners create a reference to the noun, rather than a modification thereto, so saying 'a dog' doesn't change the quality of the dog. Genitives like "Mike's" is also a determiner, for the same reason, but also a noun cannot take two determiners, so there can be "the dog" or "Mike's dog" but not Mike's the dog".

1084: Yiddish and German Nov 27, 2017

The most similar language to English is generally said to be Frisian, specifically West Frisian. Still, if an English speaker heard West Frisian for the first time, he or she not likely understand what was being said, because even though they share the Ingvaeonic root (see photo below*) there are still many centuries of foreign influence and geographic isolation that made the two languages distinct. Because they are both so famous, and because they are spoken quite near to each other, one might assume that the most similar language to Modern German would be Dutch, and while the two are certainly quite similar to the point that there is some level of mutual intelligibility (though not 100% or anywhere close to that) Yiddish is actually closer related. Though German and Yiddish occasionally differ in some vocabulary, accent, and word-order, they are remarkably similar, in part because the two only diverged relatively recently, and were spoken in a similar region.

*http://yiddish.biz/

1083: dinner Nov 26, 2017

To many people, ‘dinner’ is the last meal of the day, which can also be called 'supper' or in some places 'tea', while to many other people in certain places, 'dinner' is the mid-day meal, as opposed to something else like 'lunch'. Though 'dinner' as the final meal of the day may be more common globally, neither that nor the alternative labelling of what would otherwise be 'lunch' was not without reason. The word 'dinner' comes from the Old French, 'disner' meaning 'to dine' and when it was adopted into English, it often would denote the largest meal of the day, not when it was timed. Moreover, it is believed that 'disner' comes from 'desjëuner' meaning 'to reverse' and 'fasting', or somewhat more idiomatically: ‘to break fast’. In this way, a word that originally that either descended from or at least is related to breakfast now means 'supper' or 'lunch'.

1082: Telic and Ateilic Verbs Nov 25, 2017

There are lots of ways to describe types of words, and all of them have different applications. Words are broken up into well-known categories like ‘nouns’, ‘verbs’ ‘adjectives’ etc based upon the way they function in a clause. Beyond that, types of words in various lexical classes are also broken into categories like ‘mass nouns’ and ‘count nouns’ which have different sorts of meanings when they get pluralized or have determiners, such as ‘milk’ and ‘cookie’ respectively, but classifications for words are not always—albeit they usually are—syntactic. Telic verbs, for example, relate to actions or goals that have a defined end, such as “build” whereas atelic verbs do not. This does relate to syntax, as certain phrases can only be added to certain verbs, such as “he ran in an hour” which sounds bad, while “he built (something) in an hour” is fine, but this is also somewhat of a semantic issue.

1081: Reported Speech Overview Nov 24, 2017

There are a lot of ways to report speech, but not all of them are treated the same way, both socially and grammatically. A traditional way to do this is “[person] said/ordered/mentioned/stated/posited (that)...” which reports, obviously, what was communicated verbally. There are plenty of verbs other than “said” such as the ones before, but these all use the same syntactic structures, and while they all communicate the same thing more or less, certainly some of them would sound clunky outside of written work. Other phrases such as “[person] was like” sounds less formal generally—one wouldn’t find it in a newspaper or academic journal—and (at least for the second half of that statement) there is good reason. “[person] was like” allows the person reporting the communication to include non-verbal cues, and it is certainly possible to simple utter the phrase and make a face, or gasp etc, which wouldn’t serve written work much, but does allow for people to include much more information about past conversations. In some dialects, like MLE (spoken in London), the phrase “this is [me/them]” is used the same way as “[person] was like”.

1080: Salient Variables Nov 23, 2017

There are many different accents and dialects of English spoken all over the world, but since this is the same language, the differences will only be slight. Even if people do not know the linguistic academic terminology, people are generally able to pick out the ways variation appears, such as how the same sound can appear two distinct ways in the same context, e.g. rhotic or non-rhotic R in a New York and London accent respectively. When people are aware of this difference, it is called a 'salient variable'. Not all variables can be called salient for every person, which could be a small part of the reason why some people are bad at putting on other accents. These variables can also be emphasised or diminished in order to adapt to situations, and appear either more or less to belong to a certain group.

1079: Cooperative Principles Nov 22, 2017

Having an understandable sentiment linguistically is more than simply the grammar of a sentence when considered in isolation. The term Cooperative Principle describes the idea that people will speak in a way that is suitable to the context of the conversation, both linguistically and otherwise. The four Gricean maxims, simply put, identify rules by which people know how to talk to each other. First, there is the maxim of relevance: the notion that people will give relevant information. A somewhat exaggerated example how this is broken could be:
Speaker 1: "how's the weather?"
Speaker 2: "I like cats"
Next, there is the maxim of quality, which states that people cooperating in a conversation would tell what they believe is truthful. To lie would violate this maxim. For the following two maxims, there is some overlap. The maxim of quantity posits that people try to be as brief as they can, and the maxim of manner states that people will try to be as clear as they can—which could entail being brief—but also stating things in a way that is, for instance, chronological and grammatical. If you wanted to be annoying and confusing, you could try throwing in an irrelevant, long-winded, and unclearly stated lie into your conversation, but this also works for some jokes, such as
"In which battle did Davy Crockett die?"..."his last one", which does not violate the maxim of quality, but could be said to break the maxim of relevance, arguably.

1078: The "Do So" Test Nov 21, 2017

One way to determine when something is a verb-phrase is that it can be substituted for "do so (too)". For instance, both "walks the dog" and "walk the dog on Tuesdays" in "Yosef walks the dog on Tuesdays" are verb-phrases, which we known because one could say "...and Beth does so too (/ on Tuesday)". It is not that this test does not work for all verb-phrases exactly, but there is the problem that this works too much. In the sentence "Beth is selling her house, but Yosef could never do so", "do so" does not mean "selling her house", (or "his house", as this is not specified either) but simply "sell [his/her] house". The meaning could be essentially the same, or not, but this is not only a matter of semantics, but also one of syntax. Also, people understand that certain meaning is not carried forward with 'do so', such as "Yosef pet the cat, but he never would have done so with a dog". Since 'do so' corresponds to "pet the cat", a possible—though improbable—interpretation could be "...pet the cat with a dog" i.e. not his hand. This example relates more to semantics, but either one is an instance of the human ability to understand how—in this case—verbs are used, and have to change based on either the grammar or the literal meaning.