Morphology PDF

# 4 Morphology ## 4.1 What is Morphology - Morphology is a sub-field of linguistics that studies the internal structure of words. - It tries to describe which meaningful pieces of language can be combined to form words and what the consequences of such combinations are on the meaning or the gramma...

# 4 Morphology ## 4.1 What is Morphology - Morphology is a sub-field of linguistics that studies the internal structure of words. - It tries to describe which meaningful pieces of language can be combined to form words and what the consequences of such combinations are on the meaning or the grammatical function of the resulting word. - Every speaker of English knows that *wind* is an English word, as are *unwind*, *rewind*, *winding*, *windable*, *windy*, etc. - However, even though *woman* is also an English word, none of the following are posible: *unwoman*, *rewoman*, *womaning*, *womanable*, *womany*, etc. - Why is it that you can add *re-* to wind and get another word, but adding *re-* to woman does not result in a word? ## 4.1 Words & Words Formation: The Nature of the Lexicon - Every language has some (large) number of words available for its users to choose from as they need. - This stock of words can be thought of as a sort of mental dictionary that language users — both speakers and hearers —- have internalized as part and parcel of acquiring their particular language. We call this mental dictionary the lexicon. - But what exactly are the sorts of things we might have in our lexicon? - In the study of morphology, one topic we will consider is how words are made, but first we must answer the question of what words are. - Most everyone has an idea of what a word is. - However, not all words are equally distinct from all other words. - To begin, consider the following questions: - Are *cat* and *dog* the same word or different words? - Your answer, like that of almost anyone familiar with Einglish, is very probably: “Of course they are different words! Isn’t it obvious?” - The reasons that this is obvious include both differences in form, that is, what a word sounds like when spoken (/kæt/ is quite distinct from /dag/; refer to the chart at the end of the book for help with any unfamiliar symbols) and differences in meaning, such as the fact that you cannot simply use *cat* and *dog* interchangeably to mean the same thing. - On the other hand, you might say *cat* and *dog* are both kinds of pets, so the words aren’t 100% different; they do have something to do with each other. - These sorts of similarities, however, are not enough to lead us to claim that *cat* and *dog* are the same word. - Are *cat* and *catalog* the same word or different words? - Based on the discussion above, some readers might hesitate before answering this question. - These two words share some elements of form, the / kæ t / part, but *catalog* doesn’t seem to have the meaning of *cat* anywhere in it. - Similarly, the words *kid* and *kidney* may sound partly the same, but it seems that they are not actually related in their meaning. - Even though it sounds like there could be a *cat* and a *log* in *catalog*, or a *kid* in *kidney*, and such a conneciton might even be used as a source of humor in a joke or cartoon, English speakers consistently distinguish these pairs as each containing two unrelated words. - Thus, when looking to see whether two items are the same word, we must consider both their phonological form and their meaning. - Nevertheless, the thought that one word could be found ‘inside’ another word is an important one. - Are *cat* and *catty* ('spiteful') the same word or different words? - Here, the connection is a good bit closer than in the preceding word comparisons. - Cats have gained a reputation for sometimes being vicious fighters, and it is most probably in this context that the word *catty* came into existence as part of the English language, meaning something like ‘behaving like a cat in a certain respect’. - So the words *cat* and *catty* are similar not only in terms of their form (the /kæt/ part) but also in terms of their meaning, since both (at least potentially) engender the image of nasty fighting. - Is this enough to say that *cat* and *catty* are instances of the same word? ## 4.1.1 What are words like? - Apart from having a certain phonological form and a meaning, words also belong to lexical categories, which are also sometimes called parts of speech. - Lexical categories are classes of words that differ int how other words can be constructed out of them. - For example, if a word belongs to the lexical category verb, it is possible to add -*ing* or -*able* to it to get another word (e.g., *wind* and *drink* are verbs). - If a word belongs to the lexical category adjective, you can add -*ness* or -*est* to it to get another word (e.g., *quick* and *happy* are adjecitves. - If a word belongs to the category noun, you can usually add -*s* to it to make it plural (e.g., *desk* and *dog* are nouns. - You can add -*like* to nouns to form an adjective (*woman-like*, *city-like*, etc.) - You can also add -*ly* to many adjectives and form an adverb (*quickly*, *happily*, and *readily*). ## 4.1.2 Derivation - Nouns, verbs, adjecitves, and adverbs are also called open lexical categories because new words added to the language usually belong to these categories. - In contrast, closed lexical categories rarely acquire new members. - Closed lexical categories include: prounouns (*we*, *she*, *they*), determiners (*a*, *the*, *this*, *your*), prepositions (*on*, *of*, *under*, *for*), and conjunctions (*and*, *or*, *but*). - Now we can consider whether *cat* and *catty* belong to the same lexical category. - The answer is no - *cat* is a noun, while *catty* is an adjective. - Even though *cat* and *catty* share elements of form and elements of meaning, the fact that the words belong to different parts of speech classes is a pretty clear sign that we are in fact dealing wtih two different words, rather than two ‘versions’ of one word. - There remains the feeling, however, that *cat* and *catty* are related in a way that *cat* and *dog*, on the one hand, and *cat* and *catalog*, on the other, are not. - What is the nature of this relation? - Let’s compare some of the attribuets of the two words: | | **Cat** | **Catty** | |--------------|-----------------------------|-------------------------------------------| | **Form** | /kæt/ | /ke ti/ | | **Meaning** | ‘domesticated feline’ | ‘spiteful, (fighting) like a domesticated feline’ | | **Lexical Category** | noun | adjective | - With respect to form, *cat* is obviously a shorter word (i.e., contains fewer sounds) than *catty*. - The meaning of *catty* also seems to be based on the meaning of *cat*, rather than the other way around. - This suggests that *catty* is based on *cat*, or, in other words, that *cat* is the root on which *catty* is built. - This process of creating words out of other words is called **derivation**. ## 4.1.3 Inflection - Derivation takes one word and performs one or more “operations” on it, the result being some other word, often of a different lexical category. - When the resulting new word is not of a different lexical category, the meaning of the root word is usually changed. - For example, while *playable* is an adjective derived from the verb *play*, *replay* is a verb divided from a verb, and *unkind* is an adjective derived from an adjective, but with altered meanings: *re-* derives verbs meaning ‘Verb again’ (where “Verb” stands for the meaning of the root verb, e.g., *replay ‘play again’*), and *un-* derives adjectives meaning ‘not Adjective’ (e.g., *unkind ‘not kind’*). - In the simplest case, the root is used ‘as-is’, and one or more additional pieces are tacked onto it. - The added pieces are called affixes. - The thing to which the affixes attach is called the **stem**. - In the case of *catty*, *cat* /kæt/ is both the root and the stem, and the affix is /i/, spelled <y>, which is attached to the end of the stem. - Affixes such as /i/ are called **derivational affixes** since they participate in derivational processes. - At this point, there is one more question for you to consider: - Are *cat* and *cats* the same word or different words? - In terms of phonologial form, the difference between /kæt/ and /kæts/ is exactly the same in degree (that is, one additional sound) as the difference we saw between /kæt/ and /kæti/. - With respect to meaning, however, *cat* and *cats* seem to refer to the same kind of thing, the difference being whether we want to talk about one (singular) or more than one (plural) of that thing. - Moreover, these are both of the same lexical category, noun: | | **Cat** | **Cats** | |-----------|---------|---------| | **Form** | /kæt/ | /kæts/ | | **Meaning** | ‘domesticated feline’ | ‘domesticated feline’ (plural) | | **Lexical Category** | noun | noun | - This time the answer to the “same or different” question is not as obvious as it was in the earlier cases. - *Cats* represents a different grammatical form of the word *cat*, used just in case we need to talk about more than one member of the class of *cat*. - The creation of different grammatical forms of words is called **inflection**. - Inflection uses the same sort of pieces, such as stems and affixes, or processes that derivation does, but the important difference is the linguistic entity that inflection creates - forms of words, rather than entirely new words. - For example, in contrast to derivational affixes, inflectional affixes such as -*s* typically do not change the lexical category of the word - both *cat* and *cats* are nouns. - Similarly, both *wind* and *winding* are considered verbs. - The forms that result from inflection are often required by the grammar, regardless of any meaning difference. - For example, while *cats* means ‘more than one cat’, the plural ending is always required in English, even if the plural meaning is conveyed elsewhere: *I have four cat* is ungrammatical, even though the meaning is clear. - In contrast to the meaning differences of derived forms, e.g., *She is kind/She is unkind/She is not kind*. - In sum, we find that the idea of “same” or “different” with respect to words can be unexpectedly complicated since words have a number of differnt properties that need to be considered; at the very least, these include phonological form, meaning, and lexical category. ## 4.1.4 Some notes about morphemes - There are actually very few inflectional affixes in English, so it may help to collect them in one table for easy reference. - The following chart shows all of the functions of inflectional affixes of English and most of the common forms that those affixes take. - However, there are some less commone affixes that do not appear in the table. - For example, the plural of *ox* is formed with the suffix -*en*, but because the plural marker -*en* appears on very few words, it si not listed below. - Notice that all of the inflectional affixes in the table - and all of the inflectional affixes of English - are attached after the stem. - (Derivational affixes in English may attach either before or after the stem.) - However, this generalization does not hold for all languages. | **Function** | **Affix(es)** | **Attaches to** | **Example** | |--------------|--------------|---------------|------------------------------------------------| | 3rd per. sing. present | -*s* | verbs | She *waits* there at noon. | | past tense | -*ed* | verbs | She *waited* there yesterday. | | progressive aspect | -*ing* | verbs | She is *waiting* there now. | | past participle | -*en*, -*ed* | verbs | Jack has *eaten* the cookies. Jack has *tasted* the cookies. | | plural | -*s* | nouns | The *chairs* are in the room. | | comparative | -*er* | adjectives, adverbs | Jill is *taller* than Joe. | | superlative | -*est* | adjectives, adverbs | Ted is the *tallest* in his class. | - So far we have seen words htat cannot be broken down into smaller parts, like *cat* and *catalog*, and words that contain two parts - a root and an affix - like *catty* and *cats*. - Roots and affixes are called **morphemes**. - A morpheme is typically defines as the smallest linguistic unit with a meaning (e.g., the morpheme *cat*) or a grammatical function (e.g., the morpheme -*ed* that indicates past tense). - Of course, a morpheme also has a certain phonological form. - Thus, there are no smaller forms that carry their own meaning or grammatical function than morphemes. - A few notes are in order about the terminology that we use to discuss morphemes. - First, while a root by definition contains only one morpheme, a stem may contain more than on e morpheme. - For example, in *cattiness*, the root is *cat*, but the stem to which the derivational affix -*ness* is added is *catty*, which itself contains two morphemes, as we have already observed. - Each affix is also a single morpheme. - Affixes that follow a stem are called **suffixes**, whereas affixes that precede a stem are called **prefixes**. - Another thing to notice about affixes is that sometimes differnt meanings or functions can be marked by the same phonetc shape (note the two -*s* affixes in the previous table). - Affixes that sound alike but have different meanings or functions are **homphonous**. - (Different words that sound the same are likewise said to be homophonous.) - Another example is the case of -*er*, which can be wither inflectional or derivationl. - As an inflectional suffix, it marks comparative degree on adjectives and adverbs (like in *taller*, *faster* in the table, but the same phonetc shape can be used to derive an agent noun from a verb, as in *speak*, *speaker*. - These two -*er* affixes are homophonous with each other, and it is therefore important to consider not only form but also meaning when you are analyzing morphological structures. - Further evidence that both form and meaning are necessary when identifying morphemes comes from cases of words that merely appear to contain multiple morphemes, but infact do not. - Look again at the word *catalog*. - In terms of both its orthography and its pronunciation, it appears to contain the words *cat*, *a*, and *log*. - Neither felines nor sections of tree limbs have anything to do with ‘inventories’, though. - Thus, we conclude that *catalog* is monomorphemic: it is made of only one part. - As a final caution, do not confuse word length with number of morphemes. - Some words, such as *Madagascar*, *lugubrious*, or *pumpernickel*, are quire long but contain only one morpheme; other words, such as *ads*, are very short but contain two morphemes. ## 4.1.5 Classifying Elements In Morphology - In morphology, the most basic act of analysis is a comparison of words base on form, meaning, and lexical category. - Such comparisons allow for the segmentation of words into the smaller parts that they contain; i.e., **morphemes**. - From such an analysis, it becomes apparent that words and affixes do not share the same status in the language overall. - Simple words like *cat*, *dog*, *book*, and *walk* cannot be broken down into smaller meaningful pieces - they consist of exactly one morpheme. - Affixes like -*ing* or -*y* also consist of only one morpheme but cannot stand alone like single-morpheme words. - Morphemes such as the simple words above are called **free morphemes** because they can be used as words all by themselve. - Affixes, on the other hand, always have to be attached to the stem of some word in order to be used. - Because they cannot stand alone, affixes are called **bound morphemes**. - Affixes are not the only things that can be bound. - There are some roots that do not have stand-alone forms; that is, they only appear with one or more affixes attached. - For example, the words *infer*, *confer*, *refer*, *defer*, *prefer*, and *transfer* all seem to have a root -*fer* (stem /fגְ/( with a prefix attached to its left. - This root, however, does not correspond to any free morpheme in English. - The same is true of *boysen-* and *rasp-* in *boysenberry* and *raspberry*. - While *berry* si a free morpheme, neither *boysen-* nor *rasp-* can stand alone. - Morphemes of this sort are called **bound roots** because although they do seem to have some associated basic meaning (in the case of -*fer*, the meaning is something like ‘carry, bring’), they are unable to stand alone as words in their own right. - Other examples are -*ceive* (conceive, receive, deceive) and -*sist* (resist, desist, consist, subsist). - Can you think of a single basic meaning for each of these bound roots? ## 4.2 Morphological Processes - To this point, our morphologcial dscussion has been limited to the process of **affixation**. - Although English uses only prefixes (affixes that precede the stem they attach to) and suffixes (affixes that follow the stem they attach to), many other languages use infixes as well. - **Infixes** are inserted within the root morpheme. - Note that English has no regular infixes. - At first glance, some students think that -*ful* in a word like *doubtfully* is an infix because it occurs in the middle of a word; File 4.4 will provide a more thorough account of how affixation works and show why this must be and incorrect analysis. - In some colloquial speach or slang, there is some evidence of English infixes, but although some of these forms may be moderately productive, they are far from routinized. ## 4.2.1 The Processes of forming words - **Compounding** is a process that forms new words not by maeans of affixes but from two or more independent words. - The words that are the pats of the compound can be free morphemes, words derived by affixation, or even words formed by compounding themselves. - Examples in English of these three types are shown in the following example: | | | | |---------------------------|----------------------------------------------------|---------------------------------------------------------| | **Compounding of Free Morphemes** | **Compounding of Affixed Words** | **Compounding of Compounded Words** | | girlfriend | air-conditioner | lifeguard chair | | blackbird | ironing board | aircraft carrier | | textbook | watch-maker | life-insurance salesman | - Notice that in English, compound words are not represented consistenly in writing. - Sometimes they are written together, somstimes they are written with a hyphen, and sometimes they are written separately. - We know, however, that compounding forms words and not syntactic phrases, regardless of how the compound is written, because the stress patterns are different for compounds. - Think about how you would say the words *red neck* in each of the two following sentences: - The wool sweater gave the man a *red neck*. - If you want to make Tim really angry, call him a *redneck*. ## 4.2.2 Affixation - **Reduplication** is a process of forming new words by doubling either an entire free morpheme (total reduplication) or part of it (partial reduplication) - English makes no systematic use of reduplication as a part of the language’s grammar. - There are a very few nonsystematic cases of lexical reduplication, however, such as “*bye bye*”. - Furthermore, in colloquial speech, we may often se reduplication usd to indicate intensity; this can happen with verbs, adjecitves, and nouns. - Consider examples (11) – (13): what does the reduplicated word mean in each case? - Do you just hate him, or do you *hate-hate* him? - That shirt isn’t what I had in mind; it's much too pale of a geen. I want a shirt that is *green-green.* - Last summer we just went to Dubai, but this year we’re going on a *trip-trip*. - As you can see, though, each of these uses is very restricted to the context in which it appears. - We wouldn’t want to say that *green-green* is a word of English. - On the other hand, there are some languages that make extensive use of reduplication. - In these languages, reduplication can serve some of the same functions that affixation serves in English. - Besides adding an affix to a morpheme or copying all or part of the morpheme to make new words or make morphologiacl distinctions, it is also possible to make morpheme-internal modifications, called **alternations**. - While alternations have to do with the sounds in a particular word pair or larger word set, these alternations mark morphological distinctions, whereas the rules in the phonology files dealt with pronunciation independent of meaning. - The following are examples of morphologcal alternations in English: | | | |------|----------| | *ring* | *rang* *rung* | | *drink* | *drank* *drunk* | | *swim* | *swam* *swum* | | *feed* | *fed* *fed* | | *hold* | *held* *held* | - Although the usual pattern of plural formation is to add a suffix, some English plurals make in internal modification for this inflected form: | | | |------|----------| | *man* | *men* | | *woman* | *women* | | *goose* | *geese* | | *foot* | *feet* | - The usual pattern of past and past participle formation is to add an affix, but some verbs show an internal alternation: | | | | |--------------|--------|-----------| | *ring* | *rang* | *rung* | | *drink* | *drank* | *drunk* | | *swim* | *swam* | *swum* | | *feed* | *fed* | *fed* | | *hold* | *held* | *held* | | *break* | *broke* | *broken* | | *speak* | *spoke* | *spoken* | | *bite* | *bit* | *bitten* | | *fall* | *fell* | *fallen* | | *give* | *gave* | *given* | ## 4.2.3 Affixation in Signed languages - **Suppletion** is a completely irregular situtation where a root will have one or more inflected form phonetically unrelated to the shape of the root. - Example: - **[IZ]** *is* - **[gou]** *go* - **[wʌz]** *was* - **[went]** *went* - Interestingly, verbs derived from the irregular *go* also show similar suppletion in their past stems: *undergo*, [past] *underwent*. - Two common English adjectives—good and bad—have suppletive comarpative adn superlative forms: | | | | |---------|-----------------|---------------------------| | Adj | Comparative | Superlative | | [gud] *good* | [ber] *better* | [best] *best* | | [bæd] *bad* | [wļs] *worse* | [wļst] *worst* | ## 4.2.5 Reduplication - **Alternations** are derivational relations such as a change in aprt of speach class can be indicated by means of alternations. - For example: the final consonant of a noun voices in order to become a verb. - *strife* (n) *[straif]* - *strive* (v) *[stiaiv]* - *teeth* (n) *[ti0]* - *teethe* (v) *[tið]* - *breath* (n) *[brɛθ]* - *breathe* (v) *[brið]* - *use* (n) *[jus]* - *use* (v) *[juz]* ## 4.2.6 Alternations - Languages that employ morpholigcial processes to form words will usually have a regular, productive way of doing so according to one or more of the processes discussed above. - They might also have some smaller classes of words that are irregular because they mark the same morphologcial distinction by another of these processes. - Sometimes, however, a root will have one or more inflected forms phonetically unrelated to the shape of the root. - This completely irregular situtation is called **suppletion**. ## 4.2.7 Suppletion - In the previous file, we looked at how words are put together and marked for grammatical features such as number and tense. - We have seen that English makes use of derivational affixes to create more words than would exist with free morphemes alone. - Of course, English is not the only language that enlarges its vocabulary in this way. - When linguists observe a language that uses affixation to form additional words, they note that the occuring combinations are systematic; i.e., **rule-governed**. - Because these combinations are rule-governed, we can say that a process is at work - namely, a word formation process - since new words or forms of words are being formed. - What we will consider in this file are the ways in which languages create new words from existing words, and the grammatical forms of words. - We shall see that many languages employ affixation but that many other languages employ other processes. ## 4.3 Morphological Types of Languages - So far, we have considered a number of processes taht a language might utilize in order to form words: affixation, compounding, reduplication, alternation, and suppletion. - Some languages make use of a number of these processes; others make use of very few; still others make use of none at all. - Languages can be classified according to the way in which they use or don’t use morphological processes. - There are two basic morphological types, **analytic** and **synthetic**, the latter having several subtypes. ## 4.3.1 Classifying languages by morphological types - **Analytic languages** are so called because they are made up of seqeunces of free morphemes —- each word consitis of a single morpheme, used by itself with meaning and function intact. - Purely analytic languages, also called isolating languages, do not use affixes to composw words. - Semantic and grammatical concepts that are often expressed in other languages through the use of affixes are expessed by the use of separate words in analytic languages. - Mandarin Chinese is an example of a language that has a highly analytic structure. - In the example sentences below, for instance, the concept of plurality and the concept of the past tense are communicated in Mandarin through the use of invariant function words rather than the use of a change of form (cf. English, I to we to indicate plurality) or the use of an affix (cf. English -*ed* for past tense). - **[ พว | mən tan tçin ]** (tones omitted) - *พว* - plural - *mən* - play - *tçin* - piano - ‘We are playing the piano’ - **[ พว | mən tan tçin lǝ ]** (tones omitted) - *พว* - plural - *mən* - play - *tçin* - piano - *lǝ* - past - ‘We played the piano’ - Note that the form of ‘we’ (I-plural) that is used in the subject position is [wə mən] and that the pronoun has the same form when it is used as the object, paced after the verb: - **[ ta da WO mən ]** (tones omitted) - *ta* - s/he - *da* - hit(s) - *WO* - - *mən* - plural - ‘S/he hits us’ - Only the position of a word in a sentence shows it function. ## 4.3.2 Analytic Language - In **synthetic languages**, bound morphemes are attached to other morphemes, so a word may be made up of several meaningful elements. - The bound morphemes may add another element of meaning to the stem (derivation) or indicate the grammatical function of the stem in a sentence (inflection). - Recall that the term stem rfers to that part of the word to which affixes are added. - It may consist of one or more morphemes: for instance, in *reruns*, -*s* is added to the stem *rerun*, which is sielf made up of two morphemes: *re-* and the root *run*. - Hungarian is a synthetic language. - In the examples below, bound morphemes show the grammatical functions of nouns in their sentences: - **[ oz ɛ mber la:tjo ว kuca: t ]** -*oz* - the -*ɛmber* - man-(subject) -*la:tjo* - sees -*ว* - the -*kuca:t* - dog-(object) - ‘The man sees the dog’ - **[ ว kuco la:tjo oz em bert ]** -*ว* -the -*kuco* - dog -*la:tjo* - sees -*oz* - the -*embert* - man-(object) - ‘The dog sees the man’ - As mentioned above, in English it is the positino in the sentence of the noun phrase *the man* or *the dog* that tells one whether the phase is the subject or object of the verb, but in Hungarian, a noun phrase may appear either before or after the verb in a sentence and be recognized as the subject or object in either postion because it is marked with a bound morpheme (the suffix [t]) if it is the direct object. - (Many synthetic languages behave similarly.) - Therefore, both examples below meant the same thing, even though the position of the noun phrase meaning ‘the man’ is differnt with respect to the verb meaning ‘sees’. - **[ ว kuco la:tjo oz em bert ]** -*ว* -the -*kuco* - dog -*la:tjo* - sees -*oz* - the -*embert* - man-(objet) - ‘The dog sees the man’ - **[ oz em bert la:tjo ว kucɔ ]** -*oz* - the -*embert* - man-(object) -*la:tjo* - sees -*ว* - the -*kucɔ* - dog - ‘The dog sees the man’ - Synthetic languages like Hungarian may also use bound morphemes to indicate some concepts taht English signals by means of free morphemes. - For example, Hungarian indicates personal possession and location by the use of suffixes attached to the stem ([*haz*], ‘house’), whereas in English these concepts are expressed by the use of free morphemes. - Examples are given in (8) and (9). - **[ ว ha:zunk zøld ]** -*ว* - the -*ha:zunk* - house-our -*zøld* - green - ‘Our house is green’ - **[ ว se ked ว ha:zunkboท von ]** -*ว* - the -*se ked* - chair-your -*ว* - the -*ha:zunkboท * - house-our-in -*von* - is - ‘Your chair is in our house’ ## 4.3.3 Synthetic languages - When a linguist comes in contact with a new language, one of his or her major tasks is to discover the meaningful units that make up the language. - Just as with discovering phomemes and allophones, it is important that the linguist have procedues for discovering these minimal units, since it is impossible to isolate morphemes by intuition. - For example, the Classical Greek word *[graphɔ:]* means ‘I write’, but if the word is considered in isolation, the linguist has no way of knowing what sound or sequence of sounds corresponds to ‘I’ and which sequence corresponds to ‘write’. - In fact, the linguist has no way of knowing even whether the word can be broken down into obvious parts or whether this form was created through alternation or suppletion. - It *is* only by comaring *[grapho:]* with another form, fr instance, *[graphe:]* ‘s/he writes’, that one is able to determine what the morphemes of these Greek words are. - Looking at these two forms together allows us to hypothesize that *[graph]* is the part that means ‘write’. - Comparison, then, is the best way to begin morphological analysis. - But, of course, you will not want to compare just any forms. - Comparing a Greek word like *[phe:mi]* ‘to speak’ with *[graphɔ:]* will not provide much information, since the forms are so dissimilar and seem to have no morpheme in common. - What must be compared are partially similar forms, in which it is possible to recognize recurring units. - In this way we can identify the morphemes from which words are composed. - Let us consider our Classical Greek example once more. - If we compare *[grapho:]* with *[graphe:]* ‘he writes’, we note similarities between the forms. - The sequence *[graph-]* appears in both forms, *[graph-e:]* and *[graph-ɔ:]*, and if we compare these to the English correspondences, we find that the meaning ‘write’ appears in