levels of linguistic analysis:
how do we define language? in 1960, linguist charles hockett published "the origin of speech," detailing 13 'design features' characterizing properties of human language. in his view, all 13 properties must be demonstrated by a communication system in order to be considered a language. for example, animal communication systems have subsets of these features (depending on species) but are not considered 'languages.'
though we still use many of these design features today, a few are outdated. hockett's first feature proposed that language is "spoken and heard" through vocal-auditory channels. of course, since the 60s, we now recognize non-spoken languages (such as ASL) as fully-fledged languages.
vervet monkeys have different alarm calls for different predators (snakes, eagles, leopards) that are a combination of nature and nurture, as babies can make these calls but do not seem to understand when to properly use them.
honeybees have a 'waggle dance' to communicate the location of their food in relation to their hives and the quality/quantity of food. these dances seem to be innate, as bees from different regions do not learn each other's dances if they move hives.
sparrow songs are highly structured, with distinct notes, syllables, and phrases. songs seem to have fixed meanings (courtship, danger, etc.), are culturally learned -- individual sparrows can acquire a repertoire of songs from exposure to other birds' songs -- and even display regional dialects. however, learning seems to be innately guided, as sparrows will prefer to learn the songs of their own species.
while these examples check many of the boxes for hockett's design features, they are ultimately categorized as simply communication systems instead of languages.
can animals be taught human languages? while some apes like chimpanzees and gorillas have been taught ASL and speech (in few cases), we're currently skeptical that these primates actually understand human language. while they can produce signs, a majority of these signs are unique and new, meaningful sign combinations are rare. the most common communicative intention is requesting, and many sign combinations are altogether uninterpretable ("chase toothbrush"). researchers have also attempted to teach dogs and parrots human language, with limited success.
tl;dr -- language is a communication system, but not all communication systems are languages. while we can better understand certain properties of human language by comparing it to animal communication systems found in the wild, they do not share all design features with human language (although they may have many key features). human language is unique in its ability to refer to abstract ideas and wide ranges of ideas and topics.
branches of phonetics:
because sounds are nothing more than air molecules traveling through space, we can plot their waveforms and analyze their frequencies using spectrograms. in spectrograms, speech sounds have distinct patterns of frequency spectra called 'formants' by linguists.
articulators encapsulate parts of your mouth and throat like the tongue, upper/lower lip, upper teeth and gum ridge (alveolar ridge), hard and soft palate, uvula, pharyngeal wall, and glottis. in phonetics, we call speech sounds phones (hence phonetics).
phonetic alphabets would have a one-to-one correspondence/mapping between symbols and sounds. english does not do this, meaning its orthography -- or conventional spelling system of a language -- is not transparent.
features of a phonetic transcription system are quite particular and must abide by certain characteristics. first, each symbol must only represent one phone only, and there should only be one symbol for each sound. english violates this principle with the letter "c" because it can represent two different sounds -- [k] as in "cat" or [s] as in "cymbal." second, if two sounds can distinguish one word from another, they should be represented by different symbols that are able to capture their differences in pronunciation. english, again, violates this principle with the letters "th" (think "thy" and "thigh"). third, if two sounds are very similar, with differences arising only from their context, we should be able to represent that similarity. for example, "keep" and "cool" both contain the sound [k], but the [k] in each word is slightly different because the sounds of the words are made using a single flowing action, allowing one sound to influence a neighboring sound (this is known as co-articulation).
the ipa is a standardized representation of speech sounds in written form, with a one-to-one sound to symbol correspondence. its symbols are based on the way sounds are produced in the human vocal tract. for example, the ipa chart's columns contain labels such as "bilabial" (articulated with both lips) "labio-dental" (lower lip and upper teeth), "alveolar" (articulated with tongue against or close to upper gum ridge), "retroflex" (tongue has a flat/concave shape, articulated between gum ridge and hard palate), and "glottal" (produced by obstructing airflow in vocal tract/glottis), while its rows contain labels such as "plosive" (simple stop/occlusive consonant), "nasal" (produced with airflow through nasal cavity), and "fricative" (consonant produced by forcing air through narrow channel created by placing two articulators close together).
different languages have different inventories of sound. for example, while english shares a few consonants with mandarin, sounds like [ð] (this) and [ʒ] (measure) do not exist in mandarin. on the other hand, mandarin consonants like [ʈʂʰ] (ch) or [ɕ] (x) are unique. additionally, mandarin contains far fewer vowel sounds than english, but is a tonal language, which adds a variety of sounds that english cannot encapsulate.
consonants are speech sounds that are produced with some kind of constriction/closure somewhere in the vocal tract, and are defined by the following criteria: place (where the constriction occurs), manner (the type of constriction), and voicing (do the vocal cords vibrate?).
vowels are speech sounds that are produced without constriction using just the position of the tongue. they are defined using: height (how close is the tongue to the roof of the mouth?), backness (where is the tongue in relation to the front/back of the mouth?), roundness (whether the lips are rounded or not). monophthongs are simple vowels composed of a single configuration of the vocal tract, while complex vowels, composed of multiple vowel sounds (think "I" or "boy"), are called diphthongs.
while the previous examples of consonants and vowels are small speech bites of individual sounds, in everyday use, our phrases and sentences allow all our words to run together. this kind of speech is known as running speech or continuous speech. in running speech, the pronunciation of words may be affected by surrounding words.
suprasegmental features like length, stress, intonation, and tone go 'on top' of phones. some phones can be produced for a long time, such as the [i] sound in "beat" [bit], "bead" [bid], or "bees" [bis]. in ipa, you can denote this length by using the symbol [:]. fun fact! some languages can also use vowel length to contribute to a meaning difference (in finnish, [muta] = mud while [mu:ta] = some other, [tapan] = i kill while [tapa:n] = i meet).
pitch and intonation may also vary across the length of an utterance as we talk in relatively higher or lower pitches. for example, statements and questions in english have different pitch contours (questions typically have rising inflections while statements are falling). some languages -- like mandarin, vietnamese, or thai -- use the pitch contours of individual syllables contrastively.
syllables, while simple at first glance, have a handle of terms to describe their structure -- "onset" and "rhyme" (the latter of which is composed of a "nucleus" and "coda"). in practice, let's take the words "cat," "sing," and "star." their respective syllabic structures would be:
syllables can also be more or less stressed, represented in ipa as [ˈ] and [ˌ]. acoustic/articulatory correlates of stress can be complex in their length, loudness, and extremeness of vowels. stress patterns can be used contrastively to change the meaning of the word. in english, this is commonly done to distinguish nouns from verbs (think "record" (n.) vs "record" (v.)).
tl;dr -- speech sounds can be studied in terms of how they are produced, transmitted, and received. speech involves complex motor movements which can be classified in many ways, with features of speech that can also be added to phones (such as stress and tone). additionally, sound inventories vary widely across languages.
while phonetics is about the physical speech sounds and is used to describe all potential sounds of human language, phonology is the study of how sounds are organized within a language and how they interact with each other.
each language uses only a subset of all possible phones to use in order to make words, therefore containing a different phonemic inventory. however, not all sounds are phonemes! a phoneme refers to the smallest unit of speech that can lead to a meaningful contrast between words in a language. for example, two speech sounds are phonemes if they occur in the same position and change the meaning of a word. we can then determine which speech sounds are phonemes by searching for minimal pairs -- two or more words that are identical except for a phoneme.
quick notes:
allophones are different pronunciations of the same phoneme that do not create meaning differences (noncontrastive). there are no minimal pairs for allophones.
how can we differentiate sounds between phonemes or allophones? if the occurrence of sounds in a language allows their use to be distinguishable between the meanings of words in which they appear (contrastive distribution), they are phonemes -- we can say "these are phones of separate phonemes." if the occurrence of sounds in a language are such that they are never found within the same phonetic environment, and can be predicted to occur in specific phonetic contexts (complementary distribution), they are allophones -- we can say "these are allophones of the same phoneme."
example: to pluralize an english noun, we add the phoneme /s/. however, this phoneme is not always pronounced the same! /s/ changes sound/phone between words like "cats" [s], "dogs" [z], and "glasses" [əz]. these changes are noncontrastive and can be represented by the following rules:
we can generalize these rules instead of listing every phone in the environment by defining natural groupings that help to illustrate the overall patterns in the way that people pronounce phonemes/words. natural classes are phonemes that share features in common. we can use the rows and columns on the ipa chart to help identify these natural classes.
there are several types of phonological rules: assimilation, dissimilation, aspiration, deletion, and insertion. assimilation occurs when a phoneme becomes more similar to neighboring phonemes. for example, the english plural rule assimilates the voicing of the sound that preceded it (aka voicing assimilation). on the other hand, dissimilation occurs when a phoneme becomes less similar than neighboring phonemes. in greek, stops can become fricatives when they are followed by other stops. aspirations occur when some stop consonants are breathy/aspirated near the end of pronouncing it. voiceless stops become aspirated in english when they occur at the beginning of a stressed syllable ("pat" or "top"). deletions occur when phonemes are removed; for example, the phoneme /h/ can be deleted in unstressed syllables ("he hander her his hat" -> "he handed 'er 'is 'at"). insertion refers to when a phoneme is inserted between other phonemes, usually occurring when the other ones are very similar or hard to say them back-to-back. words like "dance" and "strength" contain an inserted voiceless stop between a nasal and a voiceless fricative.
contrast and discreteness is important for language to work. there is an infinite number of phones that we can make with our vocal tracts, and these vary continuously! languages has to change this continuous variety into a discrete set of elements -- phonemes. as speakers, we have to learn to selectively pay attention to the contrasts that matter for our native languages (phonemes) and ignore the ones that don't (allophones). this is not a simple task, however; a given phoneme can be pronounced many ways.
with so many various allophones and influences like a speaker's voice, pitch, and accent affecting the actual pronunciation of a phoneme, there is high variability in how we produce speech sounds. how is it, then, that we are able to easily perceive them?
tl;dr --
what's in a morpheme? while phonemes are sounds that distinguish meanings within a language, morphemes are groups of phonemes that have meanings. think of morphemes as the smallest unit of language that has meaning or grammatical function -- for example, "cats" has 2 morphemes (cat and -s), "unhappiness" has 3 (un- and happy and -ness), and "mississippi" has 1.
morphology, then, is the study of how words are constructed out of morphemes!
you've made it to the bottom of this page. hi. to go back to the homepage, click on the site title 𓆝 𓆟 𓆞