Morphological units

Morphological units,The word, a thorny concept

Morphological units

The word, a thorny concept.


All of us who speak, we know that we speak through words. Any of us is   able to divide what he says into independent words, which means that we know how to count how many words are in a given sentence without hesitation. But certainly they put us in a bind if we ask for a definition of the word .

Why? We would all agree that if , you come, you and I hope are words, because we are able to divide the statement if you come, I hope in these basic units. What happens is that the issue   becomes more complicated with units such as round table, twenty four , etc., of which we cannot say with the same assurance whether they are words or groups of words. At first glance, we say that round table consists of two words that can separate table of round and the result table , attach it to other adjectives: square table, blue table, etc. This raises the problem that by saying that a round table is held in the Cultural Center of the neighborhood, the two words express a single concept similar to that expressed in the word debate . The meaning of the round table is not equal to the sum of the meanings of its components: a new lexical element has been created in which both words are inextricably linked that have a new meaning. This is the same as saying that a new word has been created. If we consider them as two words, it is because we write them separately and because they maintain a characteristic accentual independence, but this set should appear in the dictionary with its new meaning.

There are many similar cases: Twenty-one is one word. But thirty- one or two hundred fifty-six , are they one or more words? What happens with the day before yesterday and its synonym before yesterday ? ¿ And with that in because I please, and so on for your information ? We could provide many more examples.

Linguistics has not yet found a valid and general definition for the word concept, but it is not surprising that this is so: it is an intuitive concept, of use, not valid for the analysis as it is proposed today; This leaves us with the curious fact that this science, linguistics, cannot define one of its basic units.

What to do then? When linguistics is reorganized and seeks new paths at the beginning of the 20th century, it finds that with the new concepts that it has to handle, there is no place for the word as we intuitively understand it: it must find a unit that encompasses both round table and blue, always, that, Miguel, ay, plain, book and bookstore , and that in addition, it is possible that you should also include in that group elements such as hetero -, – cardia , tele-, pro -, – mitir ( heterosexual, tachycardia, cable car, promise, resign ), etc. , and look for solutions. Let’s see below some of the answers you have given.

The paradox we are dealing with has its origin in the fact that the usual concept of the word conceals several different realities from each other. Let’s consider several word lists:


  1. singing(rolled)
  2. singing( singing action)
  3. singing(of a hard)


  1. have, have, have, had, had, have had.
  2. clean, clean, clean, clean.
  3. Rosa,rosam , rosae , roses, rosarum , rosis .


  1. Drink, drink,beburcio , drinker.
  2. dignified, dignified, dignified, dignified.


In (1) we have three words with different meanings and also different origins, which have in common the phonic (and graphic) form: they are different but homonymous words ; We will say of each of them that it is a grammatical word , because in the lexicon of Spanish (the lexical component, remember) each would have an entry. The case in (2) is different: each set of words has only one entry in the dictionary, and it seems logical: (2a) is a list of conjugated forms of the verb to have ; (2b) is the clean adjective flexed in gender and number, (2c.) Is the list of forms that the Latin word rosa can taketo be declined It is clear that in (2a) we have several forms of the same word , and that in each of the lists it is the first of these forms that we use as a form of appointment, that is, the form we choose as a representative of the complete list . It seems logical that all of them are part of the same lexical entry because, knowing Spanish as we know it, we know that it is automatic to obtain all these forms from the citation form. In (3) the same does not happen: although there is an obvious relationship between the four words that make up each group, all of them are different words that we use in different environments: some are verbs; others, adjectives; others, nouns and others, adverbs. They are independent words but – as we said in thecole – “from the same family”.


As we see, you have to be very careful with words. Among them they adopt very diverse relationships and the way they are formed is determined by the role they will play. In (1) there is no relationship other than the phonic between them; in (2), there is a regular flexion process , which we will see next; in (3) another process, that is given derivation . Another possible process is that which appears in a round table , similar (not identical) to the one we have in icebreakers, that of the composition . We will talk about all of them in this topic.

The morphological units lower than the word.


If it is not possible to give a definition of the intuitive concept of word because it is not functional in the linguistic description, we will have to look for the units we need in the word. Through the previous examples we have been able to suspect the existence of lower-level units whose combination creates a morphological structure that is what we call word .


These units exist, and we check this by contrasting several forms of word or several words “of the same family”. Take the examples of (2b): leaving only what they have in common, we are left clean . This part of all these words is what we traditionally called root . It is an indivisible element in smaller units with meaning: nothing shorter than clean – has meaning related or not with the idea of ​​cleaning, cleaning, etc., and is limited to being a set of phonemes that lack semantic value. What remains of those words, -o, -a, -os, -as is also divisible into two groups of units, – or and – a , which indicate male and female gender, respectively, and -s , which indicates grammatical plurality. Its significant value, as we see, is different from that of limp -. This refers to something of reality (the concept of cleanliness), while -o , – a and – s have a purely grammatical meaning: the concepts they refer to have no existence in extra linguistic reality. There are, then, units lower than the word and are, at least of two types:


A minimum units with meaning (ie, which are smaller than the word and are indivisible elements having meaning) we call monemes (in American school, morphemes ), and come in two types: the lexeme s or morphemes lexicons , which have lexical meaning , which refer to concepts of reality and morphemes to dry or grammatical morphemes , which have a purely grammatical meaning.

The word from the monem point of view :

Lexemes form words alone, or more commonly, accompanied by morphemes. In the case of a tree we have the seemingly only lexeme [1] ; in trees, trees, trees, peaks , etc., they are accompanied by morphemes. Words carrying lexemes are nouns, adjectives, verbs and most of what we call adverbs.

Morphemes can also appear in isolation (case of prepositions, conjunctions, article and auxiliary verb) or, as we have seen, being part of words. Is I speak thus of free morphemes (those who form words by themselves: see, yesterday, with, though ) and morphemes linked s (which must appear necessarily part of a word: re – ado , – eda , -ista , etc.)


So the word is a linguistic unit consisting of one or more monemes that follow a fixed order (we can not change the order of monemes in disembarking to say, eg *.. BARC – em -des- ar ) and form a block inseparable. If a word has one or more lexemes ( unconstitutional , eg), its meaning is lexical; that is, it appears in the dictionary. If, on the contrary, it lacks lexemes, its meaning is grammatical (or relational), and in the dictionary an explanation of its value and its uses will appear without being able to grant a referential value (it does not refer to any object, action or quality) . Ancient Chinese grammarians called the first words fulland the second, empty words.

The types of money.


Within bound morphemes (which is also called dependent and locked ) we must make a distinction according to their function: they serve to create different forms of the same word (wiped or , wiped os ) are called inflections and that serve to create new words from existing ones are called derivative morphemes s or affixes s , and according to their placement with respect to the lexeme they are called prefix s,  if they are placed in front: anti gas, pre fixed; infix s or interfix sif they placed between the stem and the prefix or suffix in s Anchar, POLV ar eda; and suffix s if placed after the lexeme: camin ito , lapic ero .


One issue that can pose problems is to delimit exactly what lexical meaning means from what grammatical meaning means : morphemes -o, -s, -aba , etc., are units with grammatical meaning (that is, they shape purely grammatical) concepts, but – able , – ifica r, pseudo- , etc., they seem to have a lexical meaning and ungrammatical (possibility, causation, falsehood). To explain that despite this they are grammatical morphemes we have to introduce the concept of grammaticalization [2]: Although their meanings originate from non-grammatical realities, these morphemes cannot be words by themselves. They are forms enabled to cover a grammatical need: a linked form is enabled to allow the expression of meanings that are not specific to grammar. Some languages ​​grammatize some concepts, others others: roundness is not a concept that has a morpheme in Spanish that covers it, but the Navajo does; neither intentionality (which the Eskimo does), nor the season or time (yes the Basque: es – aldia  ), nor the idea that something we refer to is not an opinion of our own, but that we reflect what another has said, whether or not we agree with our opinion (the Bulgarian does) … Curiously, the tendency to be something but without ever being completely it is grammaticalized in Spanish: it is one of the values ​​of the suffix – oide : intellectual oide . We also grammatize in Spanish the size (by means of the augmentative and diminutive suffixes), the contempt for something (with the derogatory), the concept of fruit tree ( -ero in peach tree , lemon tree , etc., -al in pear tree) and many other non-grammatical, but grammaticalized concepts. The value of these morphemes is, in short, purely grammatical, not lexical. Gramaticalización are other examples of tele – with the value of “television” in Couch Potatoes or , Euro in Euro MP , bus in bonobos , etc. They are morphemes with originally grammatical meaning that indicate gender, number, case, person, voice, time, mode, aspect, grade, and few more; that is, those that belong to the category of flexible morphemes. The morphemes from grammatization are many more and are used to create new words through the process of derivation, which we will study later. They are called derivative morphemes .

Another thing that we have to keep in mind is that the morpheme is a formally minimal unit, not semantically: a morpheme is a unit that cannot be formally divided into smaller segments even if its meaning (the concept it shapes) is decomposable into semantically more basic units : we can decompose the meaning of a rifle into several concepts whose meaning is more general than that of this word: 1. weapon – 2. of fire – 3. portable – 4. used by the infantry, but we cannot segment that word (said lexeme ) in smaller morphological units: ni fu – ni silnor any other segment in which we can divide this word means anything. The smallest unit with meaning (the monema) is rifle and not any of its parts in isolation.

With some morphemes us complicates the picture: what lexical or grammatical meaning has re common in redo, cut and resent ? Of course, the three re- do not seem to mean the same . Because of examples like this, many authors believe that the important thing when identifying a monema is not that it can be given a specific meaning, but that it can be isolated in a morphological segmentation and that it fulfills a function of lexical distinction ( refer and confer differ precisely by the presence of a different morpheme in each). In all these cases we will agree that the morpho re has appeared-, let’s give it the value we give it. It is isolated by segmentation and that could be enough to postulate its existence. We say that could be because the recourse to meaning must always have the last word: no segmental boudoir as Sal on -C- it simply because the word – or salt exists in Castilian and the other morphemes are recognizable in other units: recognize by its meaning a relationship between room and room that does not exist between salt and the last. In the morphological analysis we should take into account morphological, semantic and history factors of the language.

Morphemes, allomorphs and morphs .

Let us return to example 2a: have, have, have, had, had, have had. We will agree that in all these cases we have different forms of the word have , and that they are part of the conjugation of this verb. However, unlike what we have seen so far, these words do not share the same way lexematic : some have the lexeme TEN (have, have had, I’ve had) but there are others with tien – and tuv – . They are different forms, but all will agree Spanish speakers say they are merely variants in the same way TEN . We can postulate that this formten- , is the abstract monema of which the forms ten-, tien – and tuv – are nothing more than concrete realizations in certain phonic environments. We can compare this differentiation with what happens with phonemes, abstract phonic entities that have conditioned virtual variants that we call allophones that are made by means of concrete sounds. Following this parallel, we will say that the morpheme TEN has about alomorf you or virtual variants conditioned by morphological and phonological rules that are made by morpho s :


phonology :     phoneme   ->   allophone   ->   sound

morphology :     morpheme ->   allomorph -> morpho


In the case of we have a lexeme (or lexical morpheme) with three possible allomorphs that are made in speech through the three morphs ten-, tien -, tuv – . This also happens with grammatical morphemes: the plural morpheme can appear in the form of several allomorphs:


house / house s        armchair / armchair is       Monday / Monday ø .


In the previous examples we have seen the three morphs by which we can perform the plural morpheme (of the nouns) in Spanish: -s, -es – ø , (where ø is the absence of morph ).

With the things, it seems very clear the difference between alomorfo and morpho : we will see clearer if we think that abstract morpheme person and number appear inextricably together in a single morpho -amos- in verbal love. There is only one morph that is shared by two morphemes. If we extend the phoneme-monema and allophone-allomorph parallelism to the morphs , we can say that a morph is the concretion in a set of phonemes of one or more monemas, just as a sound is the concretion of a phoneme in speech.

Amalgam morphemes, zero morpheme and discontinuous morphemes.

We have just seen that there are times when in the same morpho two or more morphemes are mixed inseparably: in love we cannot separate in the – é a morph for the notion of “past tense”, another for that of “perfective aspect”, another for “first person” and one for “unique” and various grammatical concepts more like yes we can do it (in part) am -to- ba -s . A single morph “blanket” within itself to several morphemes. These morphs that shape various morphemes are called amalgam ( morphs ) or morphs “portemantea u ” (the French word meaning ‘coat hanger’).

We have also seen that a morpheme can be performed through the absence of morph . These morphs are often called zero morphemes or ø morphemes . This is common in Spanish, which generally has the plural morphs “full” (= no zero ), while the singular is systematically represented by zero morpheme: libroø / books. We can therefore say that a zero morpheme is the significant absence of a morpheme. That there is no morph implies the existence of the opposition with a morpheme that may appear in that position.

Another curiosity that we can find is the fact that there are morphemes represented by two separate morphs . If we take any Spanish verb forms, we realize that there are times that are identified by a final morpheme (ama ste , loves ba s, loves ra s, love would s …) and others who have a morpheme front the lexeme and another after him you love do , you’d love to do , etc. Something similar happens with certain derivational morphemes: a dorm ecer , a bast ecer , in Jaular , in ar paper , etc. that form derivatives of sleep, coarse, cage and paper . In all these examples there are not two morphemes that “collaborate” to create the word in question, but only one, divided into two parts. These types of morphemes are known as discontinuous morphemes .

The word formation systems.

Imagine the following dialogue:

Well, Paco, what are you working on now?

Well, I’m drumming drums for a company in Burgos …

That has to be tough, right?

It depends on whether the drums are easily pambables or not. A normal drumo takes about five minutes, but the pambado of some can take twenty-five or thirty.

It is clear that a normal Spanish speaker will not have understood what the conversation is about. He does not know what a drumo can be and what it is subjected to when he is pambado , but if he is explained what the words pambando and drumos mean , he will have no problem understanding the entire dialogue. This ability to recognize new words is because there is a clear regularity in the processes that serve to form new words in a language. Therefore will not need to be going to explain the meanings of pambable s, pambad or or drumo , as from pambando and drumosYou can obtain the different forms used in the dialogue with the use of word formation systems that work in Spanish. These systems are the bending , the derivation and composition :


The bending is the system by which we create new forms of a word by binding inflections a basis , which can be either a lexeme, either a core lexematic formed by more than one morpheme: from lexeme blanc – we create by flexion the word forms white, white, white, white, very white, very white, etc. If instead of taking as a lexematic core a simple lexeme, we start from the blanquead- base , in which we have the lexema blanc – plus the morphemes – ea (cf. blanquear) and the morpheme of participle, we obtain bleached, bleached, etc., following exactly the same processes. Flexion gives a class of flexive morphemes to each lexical category, and so, as we shall see soon, adjectives are given morphemes of gender, number and degree ( -o, -s, – ísim – , etc.), substantive only morphemes number, the adverbs the grade ( lejísimos ), etc., many of which are mandatory for a word can be used (not word blanc – until it does not add a morpheme gender and other number).

When talking about grammarization, we already have an impact on the fact that one language can grammatize some concepts and another, others, but perhaps the most surprising of all this is the fact that grammaticalization itself gives rise to the need for grammatical agreement with some characteristic of the external references with which these morphemes are related. Proof of this is in Spanish the fact that we have to use the female morpheme when talking about a girl and the male morpheme when referring to her brother, or that when telling something that happened yesterday we have to do it using the past forms of verb. There are languages ​​in which gender is not something that characterizes nouns,father / mother in front of brother / sister . In English, for example, a noun has no gender, and something will be masculine or feminine depending on whether its referent belongs to the masculine or feminine sex . For an Englishman it is absurd that the hand is female and the foot male, for example. We will never refer to them using he or she , but it .

Other languages ​​go further by using the so-called classifiers , which are a series of flexible morphemes that must necessarily appear as part of the word, making it “agree” with the object they designate in a series of characteristics such as the fact of being elongated, round or liquid, for example. Imagine that there was a flexible morpheme in Spanish that indicated roundness, say – ondo ; We would have to say things as two manzanondos a pelotondo or bombillondo , for example. We should not be surprised by this: remember the sister / brother opposition. Just as in Spanish, there can be no noun without a number morpheme (even if it is a zero morpheme), or a personal verb without morphemes of time, appearance, person, etc., in these languages ​​you cannot say apple without the Word have a roundness morpheme. And in addition, it must agree with the determinants and adjectives that accompany it: a small apple- shaped apple ring .

Flexive morphemes form closed paradigms. This means that the list of all the flexible morphemes of a language is (synchronously) closed to the admission of new members, unlike what happens with the list of derivative morphemes: think for example of the suffix -ata , so productive in our days: bocata, tocata, drug , etc. On the contrary, it is hard to imagine the appearance of a new plural morpheme of nouns or of appearance in verbs, for example. This is an advantage, because if we could expand the number of flexible morphemes without limits, since all of them would be mandatory for one category or another, our words would be very complicated and endless.

The derivation

The derivation is the formation of new grammatical words (not of forms of the same word, as in flexion) by means of the attachment of derivative morphemes (prefixes, infixes or suffixes) to a base: trough, garbage dump; training, exclusion, work, intellectualoid , reread, flutter, dance, and so on.

Derivative morphemes allow us to use the idea expressed by a lexical base in categories other than the original: we have the white adjective , but if we want to refer to the abstract quality of the white, we have to create a noun derived from the adjective: it is whiteness . If we want to name the action of putting something blank, we must create the derived verb whiten, etc. We jump like this from the lexical category with the lexeme under the arm and transfer it from an adjective notion to a nominal or verbal one, which represents a huge economy of lexemes (or better, lexical bases). Anyway, not always that we use the derivation we change the lexical category of the original word: if instead of using suffixes we derive by means of prefixes or infixes, we will not alter the category: paint> repaint; dance> dance , etc., as we will not do when using augmentative, diminutive and derogatory suffixes.

It is possible to move from any of the major lexical categories to the others: create adjective nouns (that is, coming from adjectives, such as whiteness ), deverbal ( bleaching ) or adverbial ( remoteness ); adjectives denominal ( TV ), deverbales ( deformable ) or adverbiales ( distant ); verbs denominal ( racketeering, alunizar ), adjetivales ( fake ) or adverbiales ( away ) adverbs de adjetivales (all those who finish in -mente and some more) and, more difficultly (?), denominational and deverbal.

Differences between flexion and derivation.


According to Ignacio Bosque (1981: 133-134), these would be the main differences between the way of acting the flexion and the derivation:

Regularity : The meaning of a word with flexive morphemes is easily predictable from the meanings of lexeme and morphemes; this does not happen in words that have derivational morphemes: If old is Old Style + female + plural , we predict that good is good + female + plural , but we can not predict from caning , where we stick + blow from that big dog it means ‘hit with a dog’, for example; if a dumpIt is the place where garbage is thrown, a hat is not the place where shadows are thrown or a truck driver can throw trucks without serious risk of his life.

It is common in languages ​​such as ours that the same morph serves to represent various morphemes: in flexion we have that -or serves in nominal flexion (that of nouns, adjectives and -some-pronouns) to indicate masculine gender, while in verbal inflection indicates first person singular present indicative active: carr or Old Style or against am or . In the derivation the same thing happens: we have already seen that the suffix -ero admits a multitude of different meanings according to the base to which it is attached; the same goes for your doublet -ario ( librarian, bank ) or with -ista : machinist, socialist, revanchist .

Universality within a class. All nouns admit plural morphemes (even if they are zero morphemes), but not all verbs admit the same morphemes: sleep, heal, give birth and labor admit the morpheme – orio with the value of ‘place where the action indicated by the verb ‘: bedroom, hospital, maternity ward, laboratory , but admit not teach, run, smoking , etc.

Derivative morphemes are not necessary to express a certain content because they always admit paraphrases: their meaning can be expressed by lexical means: small house and small house ; unnecessary and not necessary, etc. Inflectional not support this paraphrases (except, of course, the metalanguage: houses equals home + plural )

As we have already mentioned, derivative morphemes can change the lexical category of the base, while flexives always maintain it.

(6 de Bosque) Flexive morphemes are frequently demanded by the syntactic structure. The verb, the adjective, the article, etc., contain a number morpheme to fulfill the grammatical need to agree with a noun; derivative morphemes appear only due to lexical or expressive needs.

The composition.

The composition differs from the previous systems in the fact that it does not start from the union of morphemes to a base, but from the union in the same lexical unit of more than one base , or – according to the more traditional idea of ​​the union two or more words . The degree of formal and semantic component integration can range from a peak melting ( noon sordomudo, stamp, umbrella ) to the relative semantic-formal independence of the members ( floor-pilot, living room, bedroom community , etc.) . To determine this degree of fusion must take into account:

If the two components maintain their accent or if they share one for both: noon has an accent, sleeper with two.

If the plural affects only the second member (total fusion) or the first, noons , but sleeping cars .

It is true that in the composition can only join the core base character is lexemátic or, as they are also compound words but, because , etc., which have joined independent morphemes. This leads us to consider that the composition always starts from the union of several monemas, each of which can function autonomously in the language, and that would allow us to differentiate the composition more clearly from the derivation with grammaticalized morphemes. According to this criterion, the eureaucrat is a derivative and non-compound word, since there are neither * euro nor * crata as autonomous words in Spanish (but see below).

Other ways to create words.

Apart from these more or less regular systems of word formation, there are other less common ones, but they are especially interesting for the speech therapist. Let’s list some:

Lexical coinage : It is the creation of totally new words without (apparently at least) motivation in their origin: gas, kódak .

Language loan or adoption : Very common throughout the history of a language, the loan consists in the adoption and adaptation of a foreign word to the flow of a language. Current cases are, for example, software, parking , etc., even with the original spelling ( crude loan or xenism ), or football, ambigu , etc., already adapted (more or less!) To Spanish.

A special type of loan is the tracing , consisting in the translation of the terms with which a concept is designated to the “borrower” language: tracing cases are skyscrapers (from skyscraper ), mouse -in computing-, which is and while tracing metaphor shuttle space -ing. space shuttle -, re ( tro ) power -ing. feedback , etc.

Cross-linking [5] : Consists of taking only part of each of the words that enter the compound. In Spanish it is relatively rare ( bus , from automobile omnibus -lat technical ‘car colectivo’- and there. Bookmobile ‘ bus-library ‘, and many trade names: Banesto, of Ban Co ‘s locker CREDI to , Bancaya , of Ban co Viz caya ), but we have examples in other languages, such as English smog of smoke + fog , motel , of mo tor + ho tel or Russian Komsomol , of KOMmunistícheskiï Soyu z Molodezh i ‘Union of Communist Youth. ” If we consider that eurócrata is made not by the morpheme euro more – Democrat , but by the prefix Euro plus a shortened form of bureaucrat , we will have the explanation as to why this word is not made but derivative. [6]

Acronymy : It is the creation of words through the union of the initials of several other words: UN, RENFE, UFO, laser, INRI, and so on. : The curious phenomenon that can occur on acronyms bypass occurs peceros ‘PCE militants’ ucedista , laser , etc.

Morphological subtraction : It is the creation of words from the apocope of others: metro ( politano ), cole ( gio ), bike ( cleta ).

Improper qualification or derivation [7] : Use of a word belonging to a lexical category as if it belonged to another: saying in is a saying , where saying has gone from being a verb to a noun; the reds , where we have a noun from an adjective, etc. Sometimes the problem may arise as to whether the transition from one category to another is due to morphological reasons (habilitation) or syntactic means, (substantiation). On this issue, v. Forest (1990: 183-191).

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Check Also