{"id":1482,"date":"2025-01-08T19:55:33","date_gmt":"2025-01-08T17:55:33","guid":{"rendered":"https:\/\/linguistica.info\/b\/lei\/?page_id=1482"},"modified":"2025-06-29T14:08:15","modified_gmt":"2025-06-29T12:08:15","slug":"7-4-phrase-structure-rules","status":"publish","type":"page","link":"https:\/\/linguistica.info\/b\/leiwp\/toc\/7-syntax\/7-4-phrase-structure-rules\/","title":{"rendered":"7.4 Phrase structure rules"},"content":{"rendered":"<p>Phrase structure is useful in the analysis of individual sentences, for example, in order to identify structural ambiguity, but it is also useful in the analysis of whole grammars. By analyzing a large number of phrases or sentences in terms of their phrase structure, we can identify the general syntactic rules that determine how phrases and sentences are constructed in a given language. Let us look at how this could be done.<\/p>\n<h2>The noun phrase<\/h2>\n<p>The example sentences in Section <a href=\"https:\/\/linguistica.info\/b\/lei\/toc\/7-syntax\/7-3-phrases\/\">7.3<\/a> contain noun phrases with different structures. Here are some examples:<\/p>\n<div class=\"example\">\n<div class=\"number\">(1a)<\/div>\n<div class=\"sentence\"><em>Zoe<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1b)<\/div>\n<div class=\"sentence\"><em>documentaries<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1c)<\/div>\n<div class=\"sentence\"><em>a documentary<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1d)<\/div>\n<div class=\"sentence\"><em>green hair<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1e)<\/div>\n<div class=\"sentence\"><em>the yellow leaves<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1f)<\/div>\n<div class=\"sentence\"><em>her papers<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1g)<\/div>\n<div class=\"sentence\"><em>a student with green hair<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1h)<\/div>\n<div class=\"sentence\"><em>she<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(1i)<\/div>\n<div class=\"sentence\"><em>Zoe\u2019s hair<\/em><\/div>\n<\/div>\n<p>Let us look at the phrase structure of each example in turn. Example (1a) and (1b) consist of a single noun and nothing else \u2014 a proper noun in the case of (1a) and a common noun in the case of (1b). The structure of these noun phrases is shown in (2):<\/p>\n<div class=\"example\">\n<div class=\"number\">(2)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> N]<\/div>\n<\/div>\n<p>This structure is not unique to examples (1a) and (1b)\u00a0\u2014 the NP <em>Aylin<\/em> and the NP <em>sushi<\/em>, for example, have the same structure, and if we think of a combination of a first name plus a last name as a compound noun, so does the NP <em>Al Gore<\/em>. So, (2) is not just a description of specific noun phrases in specific sentences, but also a general rule for making a particular type of noun phrase in English! It tells you that you can take a noun and use it as a noun phrase.<\/p>\n<p>The same is true of the phrase structure of the example in (1c) \u2014 it consists of an article and a noun, as shown in (3):<\/p>\n<div class=\"example\">\n<div class=\"number\">(3)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> ART N]<\/div>\n<\/div>\n<p>Other noun phrases with the same structure are <em>a student<\/em>, <em>the cafeteria<\/em>, <em>the apartment<\/em>, <em>the package<\/em>, <em>the grass<\/em>, <em>the train<\/em>, <em>the tree<\/em>, <em>a garden hose<\/em> (where <em>garden hose<\/em> is a compound noun), and <em>the avocado<\/em>. Again, (3) can be thought of as a description of these particular noun phrases, or as a general rule for making such noun phrases: we can make more English noun phrases by combining articles and nouns according to the phrase structure in (3).<\/p>\n<p>The NP in (1d) consists of an adjective phrase (in this case, consisting of a single adjective) and a noun, its structure is shown in (4):<\/p>\n<div class=\"example\">\n<div class=\"number\">(4)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> AP N]<\/div>\n<\/div>\n<p>Other noun phrases with this structure are <em>renewable energy<\/em>, <em>vegan pizza<\/em>, and \u2014 with more complex adjective phrases \u2014 <em>unbelievably green hair<\/em>, and <em>more convincing arguments<\/em>. As before, we can make a large number of additional noun phrases by treating (4) as a rule for combining adjective phrases and nouns.<\/p>\n<p>So far, we have three rules for creating NPs in English. Now look at the NP in (1e). It contains an article (like that in (1c)), <em>and<\/em> an adjective, like that in (1d):<\/p>\n<div class=\"example\">\n<div class=\"number\">(5)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> ART AP N]<\/div>\n<\/div>\n<p>Another NP that has this structure is <em>a fascinating but very long documentary<\/em>, and again, we can treat (5) as a rule for creating NPs by combining an article, an adjective phrase and a noun. We have four rules now, but note that the rule in 5 is just a combination of the rules in (2), (3) and (4): an English NP consists of a noun, optionally preceded by an adjective phrase, optionally preceded by an article. Using parentheses to show optionality, we can combine all our rules into one:<\/p>\n<div class=\"example\">\n<div class=\"number\">(6)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> (ART) (AP) N]<\/div>\n<\/div>\n<div class=\"box\">\n<p>Use the rule in 6 to create different English NPs. Make sure you could represent them as a tree diagram.<\/p>\n<\/div>\n<p>Next, look at the NP in (1f) \u2014 it does not have an article, instead, it has what is sometimes called a \u2018possessive pronoun\u2019, but what in linguistics is usually called a possessive determiner. Determiners are words that provide grammatical information about a noun, such as articles like <em>a<\/em> and <em>the<\/em> (which provide information about definiteness and, in some languages, number and gender), demonstratives (<em>this, that, these, those<\/em>, pointing out specific instances), quantifiers and numerals (<em>some<\/em>, <em>many<\/em>, <em>one<\/em>, <em>seventeen<\/em>, etc., providing information about quantity), and possessive determiners (<em>my<\/em>, <em>your<\/em>, etc.), providing information about possession. In order to include all of these, we can change the rule to the one shown in (7):<\/p>\n<div class=\"example\">\n<div class=\"number\">(7)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> (DET) (AP) N]<\/div>\n<\/div>\n<p>Next, look at the verb phrase in (1g). It consists of a determiner and a noun, followed by a PP, so we can simply add an optional PP to our rule:<\/p>\n<div class=\"example\">\n<div class=\"number\">(8)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> (DET) (AP) N (PP)]<\/div>\n<\/div>\n<p>This is not a complete rule for the English NP yet. It does not, for example, include the possibility of a relative clause, as in <em>the student who has green hair<\/em>. But it&#8217;s not too bad either \u2014 most NPs that you are likely to encounter in English texts are created using this rule. Except, that is, NPs that consist of just a pronoun. For these, we need an alternative rule:<\/p>\n<div class=\"example\">\n<div class=\"number\">(9)<\/div>\n<div class=\"sentence\">[<sub>NP<\/sub> PRON]<\/div>\n<\/div>\n<p>The existence of this alternative rule explains why we can test whether a sequence of words is an NP by replacing it with a pronoun!<\/p>\n<p>Finally, look at the NP in (1i). It contains the form <em>Zoe\u2019s<\/em>, i.e, a combination of the NP <em>Zoe<\/em> and the possessive clitic we discussed in Section <a href=\"https:\/\/linguistica.info\/b\/lei\/toc\/5-morphology\/5-6-word-formation-rules\/\">5.6.<\/a> In that section, we proposed a treatment of the possessive clitic as a word-formation rule, as shown in (10a), because we argued that it produces a unit that behaves like a word \u2014 a possessive determiner. However, it is a strange word-formation rule, because the clitic attaches to a phrase rather than a word or wordstem. So, alternatively, we could represent it as a phrase-structure rule, as shown in (10b):<\/p>\n<div class=\"example\">\n<div class=\"number\">(10a)<\/div>\n<div class=\"sentence\">Form: [ X<sub>NP<\/sub> =s]<sub>DET.POSS<\/sub><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">\u00a0<\/div>\n<div class=\"sentence\">Meaning: \u2018belonging to X\u2019<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(10b)<\/div>\n<div class=\"sentence\">[<sub>DET.POSS<\/sub> NP {-s}]<\/div>\n<\/div>\n<p>We could then argue that such a unit is a particular type of phrase \u2014 a \u201cpossessive phrase\u201d, and that possessive determiners are actually a particular type of pronoun that can replace this phrase (at least that would explain the traditional label \u201cpossessive pronoun\u201d)!<\/p>\n<div class=\"box\">\n<p>Since different languages have different grammars, phrase structure rules are specific to individual languages. Think about other languages you know \u2014 do they have the same NP rule as English, or a similar one, or a completely different one? If you know a language that has a different NP rule, try to formulate this rule!<\/p>\n<\/div>\n<h2>The adjective phrase<\/h2>\n<p>We have seen two types of adjective phrases so far, shown in (11a) and (11b), but there are other types of APs, such as the ones shown in (11c) to (11f):<\/p>\n<div class=\"example\">\n<div class=\"number\">(11a)<\/div>\n<div class=\"sentence\"><em><span style=\"text-decoration: underline;\">green<\/span> hair<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(11b)<\/div>\n<div class=\"sentence\"><em><span style=\"text-decoration: underline;\">unbelievably green<\/span> hair<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(11c)<\/div>\n<div class=\"sentence\"><em>Aylin is not <span style=\"text-decoration: underline;\">afraid of Zombies<\/span><\/em>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(11d)<\/div>\n<div class=\"sentence\"><em>Aylin&#8217;s paper needs <span style=\"text-decoration: underline;\">more convincing<\/span> arguments<\/em>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(11e)<\/div>\n<div class=\"sentence\"><em>Zoe\u2019s hair is <span style=\"text-decoration: underline;\">greener than the grass<\/span><\/em>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(11f)<\/div>\n<div class=\"sentence\"><em>Zombies are not <span style=\"text-decoration: underline;\">scary enough<\/span> for Aylin<\/em>.<\/div>\n<\/div>\n<p>Obviously, adjective phrases always contain an adjective. In addition, they can optionally contain adverbs, as in (11b) and prepositional phrases, as in (11c), so the phrase structure rule for adjectives would look like this:<\/p>\n<div class=\"example\">\n<div class=\"number\">(12)<\/div>\n<div class=\"sentence\">[<sub>AP<\/sub> (ADV) ADJ (PP)]<\/div>\n<\/div>\n<p>This rule also accounts for (11d), where <em>more<\/em> is a specific type of adverb (namely one that is used to form a comparative form with adjectives that do not have a morphological comparative). It also accounts for (11e), where <em>than<\/em> functions as a preposition.<\/p>\n<p>Example (11f) is not covered by the rule: although <em>enough<\/em> seems to be used as an adverb here, it follows the adjective instead of preceding it. It is the only adverb in English that behaves like this, however, so we should treat it as an exception rather than postulating a rule that says that adverbs can follow adjectives in English!<\/p>\n<h2>The prepositional phrase<\/h2>\n<p>The example sentences in the preceding section also contained a number of prepositional phrases, for example those in (14):<\/p>\n<div class=\"example\">\n<div class=\"number\">(13a)<\/div>\n<div class=\"sentence\"><em>in the cafeteria<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(13b)<\/div>\n<div class=\"sentence\"><em>about renewable energy<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(13c)<\/div>\n<div class=\"sentence\"><em>with green hair<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(13d)<\/div>\n<div class=\"sentence\"><em>with Al Gore<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(13e)<\/div>\n<div class=\"sentence\"><em>with avocado<\/em><\/div>\n<\/div>\n<p>They all have the following phrase structure:<\/p>\n<div class=\"example\">\n<div class=\"number\">(14)<\/div>\n<div class=\"sentence\">[<sub>PP<\/sub> P NP]<\/div>\n<\/div>\n<p>This is the simplest rule we have seen so far \u2014 while NPs and APs can have many different forms, depending on which of the optional elements are used, PPs always seem to have the same form.<\/p>\n<p>However, this is not quite true. First, prepositional phrases can contain adverbs:<\/p>\n<div class=\"example\">\n<div class=\"number\">(15a)<\/div>\n<div class=\"sentence\">Zoe watched a documentary <span style=\"text-decoration: underline;\">right<\/span> in the middle of the cafeteria.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(15b)<\/div>\n<div class=\"sentence\">Zoe watched a documentary <span style=\"text-decoration: underline;\">seemingly<\/span> about renewable energy, but actually advertising cold fusion.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(15c)<\/div>\n<div class=\"sentence\">Zoe watched the documentary <span style=\"text-decoration: underline;\">just<\/span> with Aylin, as no-one else was interested.<\/div>\n<\/div>\n<p>So we have to extend the rule as follows:<\/p>\n<div class=\"example\">\n<div class=\"number\">(16)<\/div>\n<div class=\"sentence\">[<sub>PP<\/sub> (ADVP) P NP]<\/div>\n<\/div>\n<p>So what about the following examples:<\/p>\n<div class=\"example\">\n<div class=\"number\">(17a)<\/div>\n<div class=\"sentence\">Noah had never seen such a violent thunderstorm <span style=\"text-decoration: underline;\">before<\/span>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(17b)<\/div>\n<div class=\"sentence\">The tram moved <span style=\"text-decoration: underline;\">forward<\/span> very slowly.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(17c)<\/div>\n<div class=\"sentence\">Finally, they reached Noah&#8217;s stop and he got <span style=\"text-decoration: underline;\">off<\/span>.<\/div>\n<\/div>\n<p>The underlined words are called \u2018spatial adverbs\u2019 in traditional grammar. However, they do not really behave like typical adverbs \u2014 for example, they cannot modify an adjective or a preposition, as the phrase-structure rules for APs and PPs suggest, and they also cannot modify verbs, which is the typical function of an adverb. Instead, they behave like PPs \u2014 they occur in the same positions, as shown in (18):<\/p>\n<div class=\"example\">\n<div class=\"number\">(18a)<\/div>\n<div class=\"sentence\">Noah had never seen such a violent thunderstorm before this year.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(18b)<\/div>\n<div class=\"sentence\">The tram moved through the rain very slowly.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(18c)<\/div>\n<div class=\"sentence\">Finally, they reached Noah&#8217;s stop and he got off the tram.<\/div>\n<\/div>\n<p>In fact, the &#8216;spatial adverbs\u2019 are sometimes the same words as the prepositions, as in (17a) and (18a) or (17c) and (18c)! So, it makes sense to assume that these spatial adverbs <em>are<\/em> actually prepositions and that the NP in prepositional phrases is optional. Our rule for PPs would then look like this:<\/p>\n<div class=\"example\">\n<div class=\"number\">(19a)<\/div>\n<div class=\"sentence\">[<sub>PP<\/sub> (ADVP) P (NP)]<\/div>\n<div>\u00a0<\/div>\n<\/div>\n<p>In fact, PPs can also consist of a preposition and another PP, as in\u00a0<em>The sun appeared <span style=\"text-decoration: underline;\">from behind the clouds<\/span><\/em>, so a more complete rule would look like this:<\/p>\n<div class=\"example\">\n<div class=\"number\">(19b)<\/div>\n<div class=\"sentence\">[<sub>PP<\/sub> (ADVP) P (NP\/PP)]<\/div>\n<\/div>\n<h2>The verb phrase<\/h2>\n<p>Finally, the examples in the preceding sections also contained a range of different types of verb phrases, some of which are shown in (20):<\/p>\n<div class=\"example\">\n<div class=\"number\">(20a)<\/div>\n<div class=\"sentence\">Zoe <span style=\"text-decoration: underline;\">slept<\/span>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(20b)<\/div>\n<div class=\"sentence\">Zoe <span style=\"text-decoration: underline;\">watched a documentary<\/span>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(20c)<\/div>\n<div class=\"sentence\">Zoe <span style=\"text-decoration: underline;\">looked tired<\/span>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(20d)<\/div>\n<div class=\"sentence\">Zoe was <span style=\"text-decoration: underline;\">talking about renewable energy<\/span>.<\/div>\n<\/div>\n<p>We know that each of the underlined sequences is a single verb phrase \u2014 it can be replaced by <em>do the same<\/em> or <em>do so<\/em> as a whole, but not in parts:<\/p>\n<div class=\"example\">\n<div class=\"number\">(21a)<\/div>\n<div class=\"sentence\">Zoe <span style=\"text-decoration: underline;\">slept<\/span> and Aylin <span style=\"text-decoration: underline;\">did the same<\/span>.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(21b)<\/div>\n<div class=\"sentence\">Zoe <span style=\"text-decoration: underline;\">watched a documentary<\/span> and Aylin <span style=\"text-decoration: underline;\">did the same<\/span> (but not: *\u2026 and Aylin did the same a telenovela).<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(21c)<\/div>\n<div class=\"sentence\">Zoe <span style=\"text-decoration: underline;\">looked tired<\/span> and <span style=\"text-decoration: underline;\">so did Aylin<\/span> (but not: *\u2026 and so did Aylin bored).<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(21d)<\/div>\n<div class=\"sentence\">Zoe <span style=\"text-decoration: underline;\">was talking about renewable energy<\/span> and <span style=\"text-decoration: underline;\">so was<\/span> Aylin\u00a0 (but not: *\u2026and so was Aylin about magical realism).<\/div>\n<\/div>\n<p>This is going to be important later on, but for now, look at the phrase structure of the individual examples:<\/p>\n<div class=\"example\">\n<div class=\"number\">(22a)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V]<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(22b)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V NP]<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(22c)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V AP]<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(22d)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V PP]<\/div>\n<\/div>\n<p>All of them contain a verb, of course \u2014 that is why they are verb phrases. In the simplest case, this is all they contain, but in addition, they can optionally contain an NP, an AP or a PP. Using parentheses to indicate optionality, as we did before, and using a slash to indicate alternatives, we could combine these rules as follows:<\/p>\n<div class=\"example\">\n<div class=\"number\">(23)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V (NP\/AP\/PP)]<\/div>\n<\/div>\n<p>However, things are a little more complex, as the following examples show:<\/p>\n<div class=\"example\">\n<div class=\"number\">(24a)<\/div>\n<div class=\"sentence\">Aylin sent Zoe a text message.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(24b)<\/div>\n<div class=\"sentence\">Noah pushed the door open.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(24c)<\/div>\n<div class=\"sentence\">Aylin put the bottles in the recycling bin.<\/div>\n<\/div>\n<p>These are the phrase structures corresponding to these examples:<\/p>\n<div class=\"example\">\n<div class=\"number\">(25a)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V NP NP]<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(25b)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V NP AP]<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(25c)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V NP PP]<\/div>\n<\/div>\n<p>In other words, VPs can contain (a) just a verb, (b) a verb and an NP, AP or PP, or (c) a verb and an NP plus an additional NP, AP or PP, so the phrase-structure rule has to look like this:<\/p>\n<div class=\"example\">\n<div class=\"number\">(26)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V (NP) (NP\/AP\/PP)]<\/div>\n<\/div>\n<p>But things are even more complicated. Recall VPs like those in (27):<\/p>\n<div class=\"example\">\n<div class=\"number\">(27a)<\/div>\n<div class=\"sentence\">Zoe watched a documentary in the cafeteria.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(27b)<\/div>\n<div class=\"sentence\">Aylin rolled the sushi with a rolling mat.<\/div>\n<\/div>\n<p>Again, we know that the underlined sequences are VPs, because they can be replaced by <em>do the same<\/em>, but in these cases, we can also replace <em>parts<\/em> of the VPs with <em>do the same<\/em>:<\/p>\n<div class=\"example\">\n<div class=\"number\">(27a)<\/div>\n<div class=\"sentence\">Zoe watched a documentary in the cafeteria\u2026<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">\u00a0<\/div>\n<div class=\"sentence\">\u2026and Aylin did the same (i.e., she also watched a documentary in the cafeteria).<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">\u00a0<\/div>\n<div class=\"sentence\">\u2026and Aylin did the same in the library (i.e., she also watched a documentary, but not in the cafeteria).<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(27b)<\/div>\n<div class=\"sentence\">Aylin rolled the sushi with a rolling mat\u2026<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">\u00a0<\/div>\n<div class=\"sentence\">\u2026 and Noah did the same (i.e., he also rolled sushi with a rolling mat).<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">\u00a0<\/div>\n<div class=\"sentence\">\u2026 and Noah did the same with a towel (i.e., he also rolled sushi, but not with a rolling mat).<\/div>\n<\/div>\n<p>In other words, the smaller sequences <em>watched a documentary<\/em> and <em>rolled the sushi<\/em> are VPs, and so are the larger sequences. Remember that we represented this structure as shown in Figure 7.4.1.<\/p>\n<div id=\"attachment_1430\" style=\"width: 429px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1430\" class=\" wp-image-1430\" src=\"https:\/\/linguistica.info\/b\/lei\/wp-content\/uploads\/2025\/01\/tree-zoe-cafeteria-1024x686.png\" alt=\"[S [NP [N Zoe]] [VP [VP [V watched] [NP [ART a] [N documentary]]] [PP [P in] [NP [ART the] [N cafeteria]]]]]\" width=\"419\" height=\"280\" srcset=\"https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/tree-zoe-cafeteria-1024x686.png 1024w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/tree-zoe-cafeteria-300x201.png 300w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/tree-zoe-cafeteria-768x514.png 768w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/tree-zoe-cafeteria-1536x1029.png 1536w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/tree-zoe-cafeteria-624x418.png 624w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/tree-zoe-cafeteria.png 1574w\" sizes=\"auto, (max-width: 419px) 100vw, 419px\" \/><p id=\"caption-attachment-1430\" class=\"wp-caption-text\">Figure 7.4.1: Tree diagram of the sentence Zoe watched a documentary in the cafeteria<\/p><\/div>\n<p>So, a VP can also consist of a VP and a PP (one or more PPs, in fact, but lets keep things simple):<\/p>\n<div class=\"example\">\n<div class=\"number\">(28)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> VP PP]<\/div>\n<\/div>\n<p>This rule cannot be combined with the rule in (26) \u2014 we need both.<\/p>\n<h2>Auxiliaries and modals<\/h2>\n<p>You might wonder what to do about auxiliary verbs, like the one in (29a), or modal verbs, like the one in (29b):<\/p>\n<div class=\"example\">\n<div class=\"number\">(29a)<\/div>\n<div class=\"sentence\">She <span style=\"text-decoration: underline;\">is<\/span> watching the documentary.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(29b)<\/div>\n<div class=\"sentence\">She <span style=\"text-decoration: underline;\">will<\/span> make tapas.<\/div>\n<\/div>\n<p>There seem to be two possibilities: either they are part of the verb phrase, as shown in Figure 7.4.2, or they form a constituent by themselves, as shown in Figure 7.4.3.<\/p>\n<p><div id=\"attachment_1485\" style=\"width: 365px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1485\" class=\" wp-image-1485\" src=\"https:\/\/linguistica.info\/b\/lei\/wp-content\/uploads\/2025\/01\/aux-hierarchical-859x1024.png\" alt=\"[S [NP [N Zoe]] [Aux was] [VP [V talking] [PP [P about][NP[AP [Adj renewable]][N energy]]]]]\" width=\"355\" height=\"423\" srcset=\"https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-hierarchical-859x1024.png 859w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-hierarchical-252x300.png 252w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-hierarchical-768x916.png 768w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-hierarchical-1288x1536.png 1288w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-hierarchical-1717x2048.png 1717w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-hierarchical-624x744.png 624w\" sizes=\"auto, (max-width: 355px) 100vw, 355px\" \/><p id=\"caption-attachment-1485\" class=\"wp-caption-text\">Figure 7.4.2: Auxiliaries as heads of VPs<\/p><\/div> <div id=\"attachment_1486\" style=\"width: 369px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1486\" class=\" wp-image-1486\" src=\"https:\/\/linguistica.info\/b\/lei\/wp-content\/uploads\/2025\/01\/aux-flat-997x1024.png\" alt=\"[S [NP [N Zoe]] [VP [V_aux_ was] [VP [V talking] [PP [P about] [NP [AP [Adj renewable]] [N energy]]]]]]\" width=\"359\" height=\"368\" srcset=\"https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-flat-997x1024.png 997w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-flat-292x300.png 292w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-flat-768x789.png 768w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-flat-1496x1536.png 1496w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-flat-1994x2048.png 1994w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-flat-624x641.png 624w\" sizes=\"auto, (max-width: 359px) 100vw, 359px\" \/><p id=\"caption-attachment-1486\" class=\"wp-caption-text\">Figure 7.4.3: Auxiliaries as independent constituents<\/p><\/div><\/p>\n<p>Unfortunately, constituent tests do not give us a clear answer as to which of the two analyses is correct. On the one hand, Fragment and movement tests suggest that auxiliaries (and modals) do not form a constituent with the verb phrase:<\/p>\n<div class=\"example\">\n<div class=\"number\">(30a)<\/div>\n<div class=\"sentence\">What is she doing? \u2014 Watching the documentary.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(30b)<\/div>\n<div class=\"sentence\">What is she doing? \u2014 * Is watching the documentary.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(31a)<\/div>\n<div class=\"sentence\">Watching the documentary, she is.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(31b)<\/div>\n<div class=\"sentence\">* Is watching the documentary, she.<\/div>\n<\/div>\n<p>On the other hand, the deletion test shows that if we have a sequence of auxiliaries and modals followed by the verb, we can only delete sub-sequences that are contiguous to the verb, as in (31), but we cannot delete an auxiliary at the beginning or in the middle of a sequence, as in (32):<\/p>\n<div class=\"example\">\n<div class=\"number\">(32)<\/div>\n<div class=\"sentence\">Who has been watching too many telenovelas?<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(32a)<\/div>\n<div class=\"sentence\">Aylin might have been.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(32b)<\/div>\n<div class=\"sentence\">Aylin might have.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(32c)<\/div>\n<div class=\"sentence\">Aylin might.<\/div>\n<\/div>\n<p>&nbsp;<\/p>\n<div class=\"example\">\n<div class=\"number\">(33a)<\/div>\n<div class=\"sentence\">* Aylin have been.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(33b)<\/div>\n<div class=\"sentence\">* Aylin might been.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(33c)<\/div>\n<div class=\"sentence\">* Aylin been.<\/div>\n<\/div>\n<p>This is more easily explained if each auxiliary or modal forms a verb phrase containing the next one, as in Figure 7.4.4, than it is if the auxiliaries each have their own branch, as in Figure 7.4.5.<\/p>\n<p><div id=\"attachment_1487\" style=\"width: 460px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1487\" class=\" wp-image-1487\" src=\"https:\/\/linguistica.info\/b\/lei\/wp-content\/uploads\/2025\/01\/aux-sequence-hierarchical-980x1024.png\" alt=\"[S [NP [N Aylin]] [VP [V_aux_ might] [VP [V_aux_ have] [VP [V_aux_ been] [VP [V watching] [NP [AP [ADV too] [ADJ many]] [N telenovelas]]]]]]]\" width=\"450\" height=\"470\" srcset=\"https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-hierarchical-980x1024.png 980w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-hierarchical-287x300.png 287w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-hierarchical-768x802.png 768w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-hierarchical-1470x1536.png 1470w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-hierarchical-1960x2048.png 1960w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-hierarchical-624x652.png 624w\" sizes=\"auto, (max-width: 450px) 100vw, 450px\" \/><p id=\"caption-attachment-1487\" class=\"wp-caption-text\">Figure 7.4.4: A hierarchically structured sequence of auxiliaries<\/p><\/div> <div id=\"attachment_1496\" style=\"width: 417px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-1496\" class=\" wp-image-1496\" src=\"https:\/\/linguistica.info\/b\/lei\/wp-content\/uploads\/2025\/01\/aux-sequence-flat-1024x561.png\" alt=\"[S\n[NP [N Aylin]]\n[Aux might]\n[Aux have]\n[Aux been]\n[VP\n[V watching]\n[NP\n[AP [ADV too] [ADJ many]]\n[N telenovelas]\n]\n]\n]\" width=\"407\" height=\"223\" srcset=\"https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-flat-1024x561.png 1024w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-flat-300x164.png 300w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-flat-768x421.png 768w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-flat-1536x842.png 1536w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-flat-2048x1122.png 2048w, https:\/\/linguistica.info\/b\/leiwp\/wp-content\/uploads\/2025\/01\/aux-sequence-flat-624x342.png 624w\" sizes=\"auto, (max-width: 407px) 100vw, 407px\" \/><p id=\"caption-attachment-1496\" class=\"wp-caption-text\">FIgure 7.4.5: A flat sequence of auxiliaries<\/p><\/div><\/p>\n<p>There are good reasons for preferring the analysis in terms of the hierarchical structure (for example, it is much easier to describe the order in which auxiliary verbs and modals occur in such sequences), and we are going to proceed as though this is the right analysis, so we need an additional phrase structure rule for VPs:<\/p>\n<div class=\"example\">\n<div class=\"number\">(34)<\/div>\n<div class=\"sentence\">[<sub>VP<\/sub> V<sub>aux\/mod<\/sub> VP]<\/div>\n<\/div>\n<p>However, in terms of constituency structure, there are good arguments for both analyses.<\/p>\n<p>Finally, note that auxiliaries can also be used as the main verb of a sentence \u2014 in this case, they follow the phrase structure rule in (26) above:<\/p>\n<div class=\"example\">\n<div class=\"number\">(35a)<\/div>\n<div class=\"sentence\">Aylin was on a train to Spain.<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(35b)<\/div>\n<div class=\"sentence\">Zoe has green hair.<\/div>\n<\/div>\n<h2>The adverb phrase<\/h2>\n<p>Adverbs are considered to be a major word class, and as such, we would expect them to head a type of phrase called adverb(ial) phrase. We have ignored this in the preceding sections and shown adjectives directly as modifiers in verb phrases and adjective phrases \u2014 not, because adverbial phrases do not exist but because they are a bit too complex to be discussed exhaustively in an introductory textbook. But we&#8217;ll give an overview of the basics to finish of this section. First, note that what is traditionally called \u201cadverb\u201d is a very heterogeneous category of words that do not behave in a consistent way. There are at least three subcategories. First, the spatial adverbs we discussed in the context of examples (17) and (18) above and which turned out to be prepositions. Second, so-called <strong>degree adverbs<\/strong> \u2014 words that indicate intensity, like <em>very<\/em>,\u00a0<em>quite<\/em>, <em>rather<\/em>, <em>extremely<\/em> etc. And third, <strong>words related in meaning to adjectives<\/strong> from which they are usually derived by the suffix {-ly} (although there are irregular cases, like the adverb\u00a0<em>fast<\/em> from the adjective\u00a0<em>fast<\/em> or the adverb\u00a0<em>well<\/em> from the adjective\u00a0<em>good<\/em>).<\/p>\n<p>Degree adverbs can modify adjectives, as in (36a), or adverbs related to adjectives, as in (36b), but they cannot modify verbs, as (36c) shows:<\/p>\n<div class=\"example\">\n<div class=\"number\">(36a)<\/div>\n<div class=\"sentence\"><em>Zoe has <span style=\"text-decoration: underline;\">very\u00a0green<\/span> hair<\/em><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(36b)<\/div>\n<div class=\"sentence\"><i><em>Zoe was <span style=\"text-decoration: underline;\">very obviously<\/span> tired.<\/em><\/i><\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(36c)<\/div>\n<div class=\"sentence\"><strong>*<\/strong><em>Zoe listened very to Aylin. <\/em>(compare (37c))<\/div>\n<\/div>\n<p>Adverbs related to adjectives, in contrast, can modify adjectives, as in (37a), other adverbs related to adjectives (although rarely), as in (37b), and verbs, as in (37c):<\/p>\n<div class=\"example\">\n<div class=\"number\">(37a)<\/div>\n<div class=\"sentence\">\n<div class=\"sentence\"><em>Zoe has <span style=\"text-decoration: underline;\">beautifully green<\/span> hair.<\/em><\/div>\n<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(37b)<\/div>\n<div class=\"sentence\">\n<div class=\"sentence\"><em>Zoe was visibly uncomfortably<\/em> <em>cold.<\/em><\/div>\n<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(37c)<\/div>\n<div class=\"sentence\">\n<div class=\"sentence\"><em>Zoe listened intensely to Aylin.<\/em><\/div>\n<\/div>\n<\/div>\n<p>Thus, we need to posit at least two types of adverbial phrases \u2014 the one in (38a), where DEG stands for \u201cdegree adverb\u201d and the one in (38b), where ADV stands for adverbs related to adjectives:<\/p>\n<div class=\"example\">\n<div class=\"number\">(38a)<\/div>\n<div class=\"sentence\">\n<div class=\"sentence\">[<sub>DEGP<\/sub> DEG]<\/div>\n<\/div>\n<\/div>\n<div class=\"example\">\n<div class=\"number\">(38b)<\/div>\n<div class=\"sentence\">\n<div class=\"sentence\">[<sub>ADVP<\/sub> (DEG\/ADV) ADV]<\/div>\n<\/div>\n<\/div>\n<p>We need two types of phrases because they occur in different rules, and we must be able to distinguish them \u2014 (38a) and (38b) occur in APs and ADVPs, (38b) also occurs in VPs!<\/p>\n<div class=\"box\">Look at all rules given previously in this section and think about whether ADV must be replaced by the rule in (38a), the one in (38b), or both.<\/div>\n<p><span class=\"nav-previous\"><a href=\"https:\/\/linguistica.info\/b\/lei\/toc\/7-syntax\/7-3-phrases\/\" rel=\"prev\"><span class=\"meta-nav\">\u2190<\/span> Previous section<\/a><\/span> <span class=\"nav-next\"><a href=\"https:\/\/linguistica.info\/b\/lei\/toc\/7-syntax\/7-5-valency\/\" rel=\"next\">Next section <span class=\"meta-nav\">\u2192<\/span><\/a><\/span><\/p>\n<p class=\"authshp\">CC-BY-NC-SA 4.0, Written by Anatol Stefanowitsch<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Phrase structure is useful in the analysis of individual sentences, for example, in order to identify structural ambiguity, but it is also useful in the analysis of whole grammars. By analyzing a large number of phrases or sentences in terms of their phrase structure, we can identify the general syntactic rules that determine how phrases [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":1401,"menu_order":4,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-1482","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1482","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/comments?post=1482"}],"version-history":[{"count":25,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1482\/revisions"}],"predecessor-version":[{"id":2149,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1482\/revisions\/2149"}],"up":[{"embeddable":true,"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/pages\/1401"}],"wp:attachment":[{"href":"https:\/\/linguistica.info\/b\/leiwp\/wp-json\/wp\/v2\/media?parent=1482"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}