Constituent (linguistics)

src: i.ytimg.com

In syntactic analysis, constituent is a word or group of words that function as a whole in a hierarchical structure. Constituent structure analysis is associated primarily with grammatical phrase structures, although grammatical dependencies also allow sentence structure to be broken down into constituent parts. Constituent sentence structure is identified using tests for constituents . These tests manipulate some parts of the sentence and based on the results, the instructions are conveyed about the structure of sentence constituents. Many constituents are phrases. Phrases are a sequence of one or more words (in several or more theories) built around a lexical item of the head and working as a unit in a sentence. The word sequence is displayed as a phrase/constituency if it shows one or more of the behaviors discussed below.

Video Constituent (linguistics)

Test for constituents

A test for constituents is a diagnostic used to identify sentence structure. There are many tests for constituents that are commonly used to identify English sentence constituents. 15 of the most commonly used tests are listed below: 1) coordination (conjunction), 2) pro-form substitution, 3) fronting, 4) do-so -substitution 5) ipsipsis, 9) pseudoclefting, 10) pacifism, 11) negligence (removal), 12) answering the ellipsis (test question), 7) clefting, intrusion, 13) wh-fronting, 14) general substitution, 15) extracting the right node (RNR).

The order in which these 15 tests are listed here corresponds to the frequency of use, the most frequently used coordination of 15 tests and the RNR being the least used. A word of warning is generally justified when using these tests, as they often give conflicting results. The test is just a rough and ready tool used by linguists to reveal clues about the syntactic structure. Some syntacticians even organize tests on a scale of reliability, with less reliable tests being treated as useful to confirm constituents though not enough on their own. Failure to pass a single test does not mean that the test string is not a constituency, and vice versa, passing a single test does not mean the test string is a constituent. It is best to apply as many tests as possible on the given string to prove or override their status as constituents.

The 15 tests introduced, discussed, and illustrated below rely mainly on one sentence:

Drunks can delay the customer.

By limiting the introduction and discussion of tests for constituents below mainly for one sentence, it becomes possible to compare test results.

Coordinate

The coordination test assumes that only coordinated constituents, that is, join the coordinator means such as and , or , or but : The next example shows that coordination identifies words as constituents:

Drunks can delay the customer.

(a) [ Drunks ] and [bums] can delay customers.

(b) The drunk [ can ] and [will] delay the customer.

(c) Drunks can [ delay ] and [drive away] customers.

(d) Drunks can delay [ customers ] and [neighbors].

Square brackets mark coordinate coordinate structures. Based on this data, one might assume that drunkards , can , delay , and subscribers are constituents in the test sentence because these strings can be coordinated with bumbs , will , evict , and neighbors , respectively. Coordination also identifies multi-word strings as constituents:

(e) Drunks can delay [ customers ] and [neighbors].

(f) Drunks can [ extinguish customers ] and [drive out neighbors].

(g) Drunks [ can extinguish customers ] and [will expel neighbors].

This data indicates that subscriber , suspends the customer , and may delay the customer is a constituent in the test sentence.

Examples such as (a-g) are not controversial as far as many sentence structure theories are ready to see the strings tested in the sentence (a-g) as constituents. However, additional data is problematic, since they indicate that certain strings are also constituents although most of the syntactical theories do not recognize them as such, eg.

(h) The drunk [ can delay ] and [will be really annoying the customer].

(i) Drunks can [ strip off this ] and [undermine them] customers.

(j) [ Drunks can ], and [they probably will], delay the customer.

This data indicates that can delay , delay this , and Drunkers can are constituents in the test sentence. Most syntax theories reject the idea that these strings are constituents. Data such as (h-j) are sometimes addressed in terms of the right-node-raising mechanism (RNR).

The problem for the coordination test represented by example (h-j) is exacerbated when one looks beyond the test sentence, as one quickly finds that coordination indicates that the various strings are constituents that most syntactic theories do not recognize as such.

(k) Sam left [ from home on Tuesday ] and [works on Wednesday].

(l) Sam left [ from home on Tuesday on his bicycle ] and [works on Wednesday in his car].

(m) The same leaves from home on Tuesday, and from work.

String from home on Tuesday and from home on Tuesday on his bicycle is not seen as a constituent in most syntactic theories, and regarding sentences (m), it is very difficult even to distinguish how one should limit the composite of the coordinate structure. The inner coordinate structure (k-l) is sometimes characterized in terms of a constituent conjunction (NCC), and the example of coordination in sentence (m) is sometimes discussed in terms of stripping and/or gapping.

Due to the suggested difficulties with the (h-m) example, many grammarians see a skeptical coordination of its value as a test for constituents. Discussion of other tests for the constituents below reveals that this skepticism is justified, because coordination identifies more strings as constituents than other tests for constituents.

Proform substitution

Substitution, or replacement, involves replacing a test string with a corresponding proform (eg pronoun, pro verb, adjective, etc.). Substitution usually involves using a definite proform like it , he , there , here , etc. In place of phrases or clauses. If such a change produces a grammatical sentence in which the general structure has not been altered, then the test string may be a constituent:

Drunks can delay the customer.

(a) They will delay the customer. ( They = Drunks )

(b) Drunks will strip them . ( them = subscriber )

(c) Drunks will do it . ( do = turn off subscribers )

These examples show that Drunks , customers , and suspend customers in the test sentence are constituents. An important aspect of the proform test is the fact that it fails to identify most subphrasal strings as constituents, ie.

(d) * Drunks do it/it put off the customer ( do/it = will )

(e) * Drunks will do it/it from the customer ( do/it = put )

(f) * Drunks will place so/it customer ( so/it = off )

(g) * Drunks will delay them . ( them = subscribers )

These examples show that the individual words will , put , die , and the customer should not be seen as constituents. This suggestion is of course controversial, as most syntax theories assume that individual words are constituents by default. The conclusion that can be achieved based on these examples, however, is that the proform substitution uses a proform that must identify only the phrasal constituents; failed to identify sub-phrasal strings as constituents.

Topicalization (fronting)

Topicalization involves moving the test string to the front of the sentence. This is a simple motion operation. Many examples of topics seem to be only slightly accepted when taken out of context. Therefore to suggest context, examples of topicization can be preceded by ... and and capital adverbs can be added as well (eg of course ):

Drunks can delay the customer.

(a)... and subscribers , drunks can certainly delay.

(b)... and put off subscribers , drunks of course can.

These examples show that customers and suspend customers are constituents in the test sentence. Topicalization like many other tests because it only identifies phrasal constituents. When the test sequence is a sub-phrasal string, topicalization fails:

(c) *... and subscribers , drunks can certainly delay.

(d) *... and can , drunks of course delay the customer.

(e) *... and put , drunks can certainly get away from customers.

(f) *... and off , a drunkard can certainly place a customer.

(g) *... and that , a drunkard can certainly delay the customer.

These examples show that customers , can , put , die , and fail > localize test. Since these strings are all sub-phrasal, one can conclude that topicization can not identify sub-phrasal strings as constituents.

Do-so -substitution

Do-so -substitution is a test that replaces the form do it ( do it , do it , do it , do it ) into the experimental sentence for the target string. This test is widely used to investigate string structures containing verbs (because do is a verb). This test is limited in its application, though, precisely because it only applies to strings containing verbs:

Drunks can delay the customer.

(a) Drunks can do it . ( do so = turn off subscribers )

(b) Drunks do it . ( do it? can put customers off )

This example shows that extinguishing the customer is a constituent in the test sentence, while b example fails to show that can extinguish the customer is a constituent, because do so can not include the meaning of the capital verb can . To illustrate more fully how to do it test is used, another test sentence is now used, which contains two post-verbal additional phrases:

We met them at the pub because we had time.

(c) We do it in the pub because we have time. ( do so = meet them )

(d) We do it because we have time. ( do it meet them in pub )

(e) We do it . ( do it = meet them in pub because we have time )

This data shows that meet them , meet them in pub , and meet them in the pub because we have time is a constituent in the test sentence. Together, such examples seem to motivate structures for test phrases that have the phrase verbs of left-branching phrases, since only the left-leaning verb phrase can see each string represented as a constituent. There is a problem with this kind of reason, however, as the following example illustrates:

(f) We do it in the pub. ( do it meet them because we have time )

In this case, doing it seems to stand for a combination of disconnected words that consist of meeting them and because we have time . The combination of intermittent words can not be construed as constituents. That such an interpretation is so indeed may be seen in fuller sentences like you meet them in a cafe because you have time, and we do it in pubs . In this case, the preferred reading do it is that it does simultaneously stand for both meet them and because we have time .

One -substitution

The one -substitution test replaces the test string with unlimited pronoun one or that . If the result is acceptable, then the test string is considered as a constituent. Since one is a type of pronoun, one -ubstitutions are only valuable when investigating the nouns noun structure. In this case, the above test sentences are extended to better illustrate the way in which one substitution is generally used:

Drunks will delay loyal customers around here that we rely on.

(a) Drunks will delay the faithful around here that we rely on. (ones = subscriber )

(b) Drunks will delay the around here which we rely on. ( ones = loyal customers )

(c) Drunks will delay the faithful whom we rely on. ( ones = customers around here )

(d) Drunks will strip who whom we rely on. ( ones = loyal customers around here )

(e) Drunks will delay the faithful who . ( ones = customers around here that we rely on )

These examples show that customers , loyal customers , customers around here , loyal customers around here , and the customers around here that we rely on are constituents in the test sentence. Some people point to issues related to one -substitution in this area, however. The problem is that it is not possible to produce a single constituent structure of the noun phrase loyal customers around here that we rely on that can simultaneously see all the constituent strings shown. Another problem that has been raised about one -sustitution as a test for constituents is the fact that at the time it indicated that the combination of non-string words is a constituent, ie.

(f) Drunks will delay the around here. ( ones = loyal customers we rely on )

The combination of words consisting of both loyal customers and we rely on is disconnected in the test sentence, a fact that should motivate a person to generally question the value of one - substitutes as a test for constituents.

Answer fragments (answer ellipsis, test question, self-test)

The answer fragment test involves the formation of a question containing one word-wh (ie who , what , where , etc.). If a test string can appear itself as an answer to such a question, then the probability of a constituent in the test sentence:

Drunks can delay the customer.

(a) Who can delay the customer? - Drunks .

(b) Who can get drunk? - Subscribers .

(c) What will a drunkard do? - Uninstall subscriber .

These examples show that Drunks , customers , and suspending customers are constituents in the test sentence. The answer fragment test is like most other tests for constituents because it does not identify sub-phrasal strings as constituents:

(d) What about firing a customer? - * Can .

(e) What can a drunkard do about a customer? - * Put .

(f) * What can drinkers do about placing customers? - * Off .

(g) * Who can drink it? - * Subscribers .

This answer fragment is all grammatically unacceptable, showing that can , put , die , and subscriber not a constituent. Note also that the last two questions themselves are not tempting. It seems impossible to form a question in a way that can successfully bring up a string that is indicated as an answer fragment. The conclusion, then, is that the answer fragment test is like most other tests because it fails to identify the sub-phrasal string as a constituent.

Clefting

Clefting involves placing an X test string in a structure that begins with This is/: That X is... . The test string appears as a pivot of the blockage phrase:

Drunks can delay the customer.

(a) This is a drunk that can quell customers.

(b) This is a drunk customer can delay.

(c) Ã, ?? This is a turn off subscribers that drunks can do.

These examples show that Drunks and customers are constituents in the test sentence. Example c is a dubious acceptance, indicating that extinguishing customers may not be a constituent in the test string. Clefting is like most other tests for constituents because it fails to identify individual words as constituents:

(d) * It can that the drunkard extinguished the customer.

(e) * This is put that the drunkard can get out of the customer.

(f) * This is off that a drunkard can place a customer.

(g) * It is that that a drunkard can delay a customer.

(h) * This is a customer that a drunkard can delay.

The examples show that each individual word can , put , die , that , and < i> the customer is not a constituent, contrary to what most syntactical theories assume. In this case, clefting is like many other tests for constituents because it only manages to identify certain phrasal strings as constituents.

VP-ellipsis (elliptical verb phrase)

The VP-ellipsis test checks to see which strings contain one or more predictive elements (usually verbs) can be circumvented from a sentence. Strings that can be interpreted are considered constituents: Symbols? used in the following example to mark the ellipsis position:

Beggar can immediately postpone customers when they arrive, and

(a) * drunks can also be immediate? customers when they arrive. (? = release )

(b) Ã,? drinkers can be soon too? when they arrived. (? = suspend subscribers )

(c) can a drunkard also? when they arrived. (? = immediately turn off subscribers )

(d) drinking soon? (? = delaying customers when they arrive )

(e) can a drunkard also? (? = immediately turn off subscribers when they arrive )

These examples show that delay is not a constituent in the test sentence, but it immediately extinguishes the customer , suspends the customer when they arrive >, and immediately turn off the customer when they arrive is a constituent. Regarding the string delaying the customer in (b), the marginal acceptance makes it difficult to draw conclusions about putting off customers .

There are various difficulties associated with this test. The first is that it can identify too many constituencies, as in this case here where it is impossible to produce a single constituent structure that can simultaneously see each of three acceptable examples (c-e) as having left a constituent. Another problem is that this test can sometimes indicate that the word combination is disconnected is a constituent, for example:

(f) Frank will help tomorrow at the office, and Susan will? today. (? = help... in the office )

In this case, it appears as if the material eluted corresponds to a combination of disconnected words including help and in the office .

Pseudoclefting

Pseudoclefting is similar to clefting that places emphasis on a particular phrase in a sentence. There are two variants of the pseudocleft test. One variant inserts an X test string in a sentence that starts with the free clause: What..... is/is X ; Another variant inserts X at the beginning of the sentence followed by it/are and then the free clause: X is/what/who... Only the two of these two variants are illustrated here.

Drunks will delay the customer.

(a) Drunk is who can delay the customer.

(b) Subscribers are the ones who get drunk can delay.

(c) Remove the customer what a drunk can do.

These examples show that Drunks , customers , and suspending customers are constituents in the test sentence. Pseudoclefting fails to identify the most individual words as constituents:

(d) * Can is a drunkard who fires a customer.

(e) * Put what can be drunk from the customer.

(f) * Off is a drunkard who can place customers.

(g) * What is that the hangover can delay the customer.

(h) * Customer is a drunk person who can place it.

The pseudoclefing test is therefore like most other tests as far as identifying phrasal strings as constituents, but it does not indicate that sub-phrasal strings are constituents.

Passivation

Passifice involves turning active sentences into passive sentences, or vice versa. Active sentence objects are converted into the subject of appropriate passive sentences:

(a) Drunk may delay subscribers .

(b) Customers may be suspended by drunk .

The fact that sentence (b), passive sentence, is acceptable, indicates that the drunk and the customer is a constituent in the sentence (a). The pacific test used in this way is only able to identify the subject and words of objects, phrases, and clauses as constituents. It does not help identify phrasal strings or other subfras as constituents. In this case, the value of pacification as a test for constituents is very limited.

Failure (deletion)

Abandonment checks whether the target string can be omitted without affecting the grammar of the sentence. In many cases, local and temporal adverbs, attributive modifiers, and optional appendages can be safely removed and thus qualify as constituents.

Drunks can delay the customer.

(a) Drunks may delay the customer. ( the has been omitted.)

This sentence shows that the definite article the is a constituent in the test sentence. Regarding the test sentence, however, the negligence test is very limited in its ability to identify constituents, since the string to be examined does not appear optionally. Therefore, a customized test sentence is used in addition to better describe the negligence test:

The obnoxious drunkard can immediately turn off the customer when they arrive .

(b) The drunkards can immediately put out the customer when they arrive. ( annoying has been successfully removed.)

(c) Irritating drunks can delay customers when they arrive. ( soon has been successfully removed.)

(d) Irritating drunks can quell subscribers. ( when they arrive have been successfully deleted.)

The ability to remove obnoxious , immediately , and when they arrive suggest that this string is a constituent in the test sentence. The omission used in this way is a limited application, since it is unable to identify any constituents that arise mandatory. Therefore there are many string targets that most accounts of sentence structure are taken to be constituents but fail in the test of omission because these constituents appear mandatory, such as the subject phrase.

Intrusion

Intrusion probes the sentence structure by having an annoying "annoying" word into the sentence section. The idea is that the strings on either side of the adverb are constituents.

Drunks can delay the customer.

(a) Drunks definitely may delay the customer.

(b) The drunkard may definitely delay the customer.

(c) * Drunks can get rid of customer definitely .

(d) * The drunkard may delay definitely the customer.

(e) * Drunks can quell customers definitely .

Example (a) shows that Drunkers and can delay a customer is a constituent. Example (b) indicates that Drunkers can and suspend customers are constituents. The combination of (a) and (b) suggests otherwise can is a constituent. Sentence (c) indicates that Drunks can get rid of and customers are not constituents. Example (d) indicates that Drunks may delay and subscribers not constituents. And exmaple (e) shows that Drunks can extinguish and customers instead of constituents.

Those who use intrusion tests usually use capital information such as definitely . This aspect of the test is problematic, since the test results may vary based on the choice of adverbs. For example, adverbs are different ways of distributing capital adverbs and will therefore suggest structural structures that are different from those suggested by capital adverbs.

Wh-fronting

Wh-fronting checks to see if the test string can be fronted as a wh-word. This test is similar to an answer fragment test so far using only the first half of the test, regardless of the potential answer to that question.

Drunks will delay the customer.

(a) Who will delay the customer? ( Who <-> Drunks )

(b) Who will get drunk? ( Who <-> subscribers )

(c) What will the drinker do? ( What... do <-> delay the customer )

These examples show that Drunks , customers , and suspending customers are constituents in the test sentence. Wh-fronting is like a number of other tests because it fails to identify many subphrasal strings as constituents:

(d) * Do what drunk fired a customer? ( Do what <-> will )

(e) * Do what drunks will be out of customers? ( Do what <-> insert )

(f) * What will customers drink? ( What <-> die )

(g) * What will get intoxicated customers? ( What <-> that )

(h) * Who will put out the drunks? ( Who <-> subscribers )

These examples show a lack of evidence to see individual words be , put , dead , and < i> subscribers as constituents.

General replacement

Adding the right node (RNR)

Maps Constituent (linguistics)

Constituency test and disambiguation

Syntactic ambiguity characterizes sentences that can be interpreted in different ways depending solely on how one perceives the syntactic relationship between words and compiles them into phrases. Possible Interpretations of the Sentence They kill the man with the gun are:

'The man was shot.'

'The murdered man has a gun with him.'

The ambiguity of this sentence results from two possible constituent arrangements:

They killed [the man] [with a gun].

They kill [man with gun].

In the first sentence, with pistol is an independent constituent with instrumental meaning. In the second sentence, it is embedded in the noun phrase man with pistol and modifies the noun man . Autonomy unit with pistol in the first interpretation can be tested by ellipsis test answers:

How did they kill the man? - With a gun .

However, the same test can be used to prove that the man with the gun in the second sentence should be treated as a unit:

Who (m) did they kill? - Men with gun .

The ability of constituency tests to distinguish certain sentences in this way bears witness to their usefulness. Most if not all syntax uses constituency tests in some form or another to arrive at the structure they set for the sentence.

src: slideplayer.com

Competing theories

The alternative theoretical approach to syntax makes different assumptions about what is considered a constituent. In the grammar of the structure of the main phrase (and its derivatives), individual words are constituents within and of themselves as well as being part of another constituent, whereas in dependent grammar, certain core words in each phrase are not constituents by themselves, but only members of the phrasal constituents. The following trees show the same sentence in two different theoretical representations, with the representation of the phrase structure on the left and the dependency grammatical representation on the right. In both trees, constituents are understood as all trees or every labeled subtree (a node plus all vertices dominated by the knot); note that words like kill and with , for example, form subtrees (and are considered constituents) in the representation of phrase structures but not in the representation of dependency structures.

src: i.ytimg.com

See also

• Catena (linguistic)
• Verbs are limited
• Unlimited verb

src: 2.bp.blogspot.com

Note

src: i.ytimg.com

References

Source of the article : Wikipedia