Abstract
In this paper, it is argued that synthetic compounds based on the passive participle in English fall into two classes, depending on whether they possess the capacity to license modifiers pointing to the presence of the external argument. Compounds such as computer-generated, pencil-drawn or home-made are typically used as eventive and resultative participles, both of which are syntactically agentive. Compounds belonging to this group are contrasted with adjectives such as action-packed, family-oriented, work-related or adult-themed, which are shown to behave syntactically like underived adjectives, with no traces of the external argument. As such, they correspond to the class of stative participles.
1 Introduction
In the morphosyntactic literature, there exist two approaches to the issue of the presence of the external argument in adjectival passives. Early accounts (e.g. Wasow 1977; Levin & Rappaport 1986) argue that adjectival passives are constructed in the lexicon (as opposed to verbal passives which are constructed in the syntax) and as such do not project the external argument. This view has been rejected by syntactic approaches to word formation, according to which the external argument is present in the structure of adjectival passives (e.g. McIntyre 2013; Alexiadou et al. 2014; Bruening 2014; Alexiadou et al. 2015). The objective of this study is to shed new light on the issue of agentivity in adjectival passives by examining the external syntax of English synthetic compounds based on the passive participle (e.g. student-written, computer-generated, home-made, action-packed). 1 Specifically, the research aims to answer the question of whether synthetic -en compounds can function as stative, resultative and eventive participles, the division proposed by Embick (2003, 2004), and whether the syntactic representations proposed by Embick can be applied to synthetic -en compounds.
The present study is performed within the Distributed Morphology framework (Halle & Marantz 1993; Marantz 1997; Harley & Noyer 1999; Embick & Noyer 2007; Embick 2010). The DM approach posits that the formation of words is driven by the same mechanism as the creation of sentences, one of the consequences of which is that the deverbal status of a (complex) word is conditioned on its external syntax.
The organization of the paper is as follows: Section 2 outlines Embick's division of participles into stative, resultative and eventive. Section 3 provides a description of the external syntax of synthetic -en compounds. Specifically, the relevant syntactic tests will be used to determine whether synthetic -en compounds are eventive and what their syntactic derivation is. The paper is concluded in Section 4.
Unless stated otherwise, the linguistic examples used in this paper have been retrieved from publications indexed in the Google books database.
2 Embick's participles
Embick (2003, 2004) divides participles in English into “stative”, “resultative” and “eventive”, as shown below with the root √OPEN (Embick 2003, 148).
| a. | The door is open. | (stative) |
| (= The door is in an open state.) | ||
| b. | The door is opened. | (resultative) |
| (= The door is in a state of having become opened.) | ||
| c. | The door was opened by John. | (eventive) |
| (= John opened the door.) | ||
Embick proposes the following structures for the three classes of participles (Embick 2003, 148–149):
|
|
|
The structure in (2a) is that of a simple adjective and as such does not contain the vP head as open does not imply that a prior event has taken place (open can be a complement of a verb of creation, e.g. The door was built open 2 ). The eventivizing vP head is present in the structure of resultative participle, indicating that the state of being opened is a result of a previous event. Crucially, the morphosyntactic feature [AG] is not represented by the vP head in the structure of the resultative participle due to the ungrammaticality of sentences such as *The door remained opened by John (Embick 2004, 357). This feature is present in the structure of the eventive passive, ensuring the agentive interpretation of the participle.
3 Synthetic -en compounds
3.1 Deverbal -en compounds
A number of deverbal -en compounds incorporate left-most elements that modify the event expressed by the verb from which the compound head is derived. Left-most elements in synthetic -en compounds typically stand in an agentive (e.g. man-made, student-written, court-approved), 3 instrumental (e.g. computer-generated, pencil-drawn, radio-controlled) or locative (e.g. home-made, pan-fried, container-grown) relationship with the head participle. As illustrated in (3), synthetic -en compounds can function as eventive participles: 4
| a. | Part of the landscape was cloaked in forest, but in the savanna Denevan saw remnants of vast earthworks so geometrical they could only have been man-made. 5 |
| b. | Agreements which have not been court approved are not enforceable and do not produce any effect (…). |
| c. | Figures 17 and 95 were kindly provided by Mrs. Marie Wurdack. All the others (…) were pencil-drawn by the author. |
| d. | The integral holograph was made from a series of two-dimensional images that were computer-generated by Stephen R. Levine (…). |
| e. | The device was homemade by an insider who was a self-taught electrician. |
| f. | Round steaks were pan fried by approximately one third of the respondents. |
Deverbal -en compounds frequently behave like adjectives. As evidenced by the examples below, deverbal -en compounds can co-occur with copular verbs, which has been recognized to be a highly adjectival feature (Wasow 1977; Levin & Rappaport 1986). In (4), the compounds contain agents, causers or instruments as their left-most elements.
| a. | There is almost nothing about the land depicted in this map that looks man-made (…). |
| b. | Annual reports cover all deaths including those that are doctor certified and those that were referred to a coroner. 6 |
| c. | (…) everything looked sun-bleached and pastel with normal-looking people walking the streets (…). |
| d. | The area looks war-devastated with blocks of gutted-out apartment buildings and boarded-up storefronts. 7 |
| e. | Many residential side streets remain snow-covered after winter storm. (article headline) |
| f. | Don't Starve is a strange beast, mixing fantasy, survival horror, construction game, exploration and elegant animation that looks pencil drawn. 8 |
| g. | There seems nothing – including print publishing – that is not computer-generated. |
Synthetic -en compounds in which left-most elements are locations are also found in adjectival contexts with Voice-related modifiers.
| a. | (…) all the food, especially the baklava and lentil soup, appeared to be homemade by someone who'd inherited generations of Middle Eastern traditional culinary skills. |
| b. | All our posters are field collected by Ernie Wolfe III. 9 |
The data in (3), (4) and (5) point to the conclusion that the external argument is present in the structure of both eventive and resultative (adjectival) variants of deverbal -en compounds. The tree in (6) is the syntactic representation of deverbal -en compounds when used as adjectives:
|
The representation is headed by aP which categorizes the structure as a syntactic adjective. The function of the PASS head is to introduce the passivizing exponent. 10 Following Kratzer (1996), we assume that the external argument is introduced in the VoiceP projection, 11 the presence of which is justified by the agentive nature of deverbal -en compounds. The structure of synthetic -en compounds used as eventive passives is identical to (6), except the lack of the adjectivizing head.
Deverbal -en compounds that denote results of an event differ from Embick's resultative participle in that they are interpreted agentively. According to Embick, the reason why resultative participles are not agentive is that the vP projected by them has an inchoative reading. Resultative adjectives such as opened can be derived only from verbs without an obligatory external argument, and not from bi-argumental verbs: 12
| a. | The vase seems broken. |
| b. | *These clothes seem made. |
| c. | *The picture seems drawn. |
| d. | *The image seems generated. |
The incorporation of the modifier into the vP domain enables participles such as made, drawn or generated to function as adjectives: 13
| a. | This garment looks to be *(home)-made. |
| b. | The lines look to be *(hand)-drawn. |
| c. | The image looks to be *(computer)-generated. |
The adjectival nature of resultative -en compounds stems also from the fact that they can modify nouns serving as arguments of verbs of creation. While, as shown by Embick (2004, 357), placing a participle such as opened in such context results in a contradiction (*The door was built opened), the same is not true for synthetic -en compounds:
| a. | Because you can't make student-written guides without the students, we have students at each campus who help write, randomly survey their peers, edit, layout, and perform accuracy checks on every book that we publish. |
| b. | My particular favorite happens to be from a pop-culture Internet site whose author has created a pencil drawn comical dragon (…). |
| c. | Hoover has produced a radio-controlled vacuum cleaner, but a spokesman denied that the company was taking robotics seriously. |
| d. | Owners may choose to prepare a homemade diet for their pet for a variety of reasons. |
| e. | You can use this same recipe to make pan-fried quail by substituting 4 whole quail, cut in half, lengthwise, for the chicken in the recipe. |
The indication of the examples in (9) is that synthetic -en compounds, despite their event semantics, are often employed in a way similar to how canonical, underived adjectives are used, which further leads to their structure being headed by aP. 14
3.2 Stative -en compounds
English has a class of adjectives composed by means of attaching a noun to the passive participle of a verb. Such adjectival compounds can nevertheless be hardly seen as expressing results of prior events (i.e they are semantically stative). Compounds of this type include action-packed, family-oriented, work-related, adult-themed, heart-shaped or home-based. The main factor pointing to the non-eventivity of such adjectives is their inability to function as eventive passives (despite their head participles being possible as verbal passives). 15
| a. | *The film was action-packed by the screenwriter. (cf. The items were packed by store employees.) |
| b. | *This country's economy was export-oriented by the government. (cf. The telescope was carefully oriented towards the Moon by the astronomer.) |
| c. | *John's injury has been work-related by the doctor). (cf. John's condition was related to the working conditions by the doctor.) |
| d. | *This book was adult-themed by the author. (cf. The Costume Institute cares for a closet of over 80,000 outfits and accessories, and seasonal shows are themed by designers or historical importance.) |
| e. | *The boxes were heart-shaped by the manufacturer. (cf. It is an undecided forest, its luxuriant ecotone having been shaped by the forces of glaciation during the Pleistocene era.) |
| f. | *This business was home-based by the owner. (cf. Their conclusion as to whether or not a given report was reproduced was deliberately based on the last experiment.) |
Compounds of the action-packed type also disallow the licensing of agent-oriented modifiers, even though the participle they are headed by can co-occur with them:
| a. | This film looks (*carefully) action-packed. (cf. What irritated him more was that his carefully packed luggage had been misplaced…) |
| b. | This community looks (*carefully) family-oriented. (cf. The grid is carefully oriented with respect to specimen axes…) |
| c. | This disease looks (*carefully) smoking-related. (cf. The social philosophy needs to embody economic principles carefully related to the various values…) |
| d. | This book is (*carefully) adult-themed. (cf. At a Disney resort, you will find yourselves the pampered guests at accommodations that have been carefully themed to enfold you in the very Disney magic…) |
| e. | The box looks (*carefully) heart-shaped. (cf. The lawn seems carefully shaped.) |
| f. | This business is (*carefully) home-based. (cf. Interventions used in clinical, behavioral, and mental health practice settings are often not carefully based on empirical evidence…) |
As also shown in (12) and (13), stative -en compounds, in contrast to deverbal -en compounds, exhibit strongly adjectival features such as the unrestricted ability to license very (see e.g. Kennedy & McNally 1999) 16 and serve as the base for -ness suffixation. Both of these properties are not exhibited by the great majority of eventive -en compounds.
| a. | For anyone who is new to Shakespeare, probably the most easily understood tragedy is Macbeth, which is short and very action-packed. |
| b. | David has always been very family oriented. |
| c. | (…) the Department is aware that their officers work under very stress-related conditions (…). |
| d. | In 1968, AIP had teamed with another Hammer rival, Tigon Pictures, for a controversial and very adult-themed film called Witchfinder General. |
| e. | They are designed for stud-necked horses that have very round-shaped necks. |
| f. | Throughout the period, Britain was a very home-based society and that characteristic grew as the quality of housing improved. |
| a. | I am excited for the season to get into the action-packedness. |
| b. | One indication of family-orientedness is the stability of marriages. |
| c. | High life-strain (…) plus inherent emotional arousal and physiologic reactive propensities define the stress-relatedness of these conditions. |
| d. | The political neutrality and single-themedness stood in contrast to the Gay Liberation Front (…). 17 |
| e. | (…) a correlation computation can't see the heart-shapedness (cardiomorphism?) of this scatterplot any more than your camera can detect gamma rays. |
| f. | Courses within each group were further sampled on a number of other criteria including course size, institutional type, geographical location, degree of ‘school basedness’ (…). |
| a. | *very student-written/court-approved/doctor-certified/pencil-drawn/pan-fried |
| b. | *student-writtenness/court-approvedness/doctor-certifiedness/pencil-drawnness computer-generatedness/pan-friedness |
Very and -ness are accepted only by compounds which express very characteristic states, such as home-made or computer-generated:
| a. | It was apparently a very homemade effort with two guys under a shoulder board (…). |
| b. | (…) many who work on computers produce work that looks very computer generated. |
Despite the presence of overt passive morphology and the compound status, stative -en compounds of the action-packed type appear to correspond to the class of stative participles. The non-eventive syntactic behavior of compounds of the action-packed type indicates that they have a structure without the vP and VoiceP layers, with the adjectivizing suffix attached directly to the root, as is the case with the structure of the stative participle.
The question which appears at this point is whether the modifier in compounds of the action-packed type is introduced by aP or by the root itself. 20 It is important to note here that left-most elements in stative -en compounds are obligatory elements and as such may be considered arguments of the head participle.
| a. | This film is *(action)-packed. |
| b. | John's condition is *(stress)-related. |
Harley (2009) assumes that it is the root that is responsible for argument selection. Harley's contention is that because argument-introducing of-phrases licensed by nominals such as student cannot be stranded under one-replacement, its arguments are introduced by the root.
| *That student of chemistry and this one of physics sit together. |
The anaphoric so, which is known to stand in for adjectives in English, differs from the anaphoric one in that (at least in contexts where the adjective and the anaphor describe the same subject) it allows the stranding of PPs licensed by adjectives.
| a. | Emerald is very happy with the book's sales, but less so with the critical acclaim. |
| b. | Amphibians from warm environments tend to be more tolerant of high temperatures and less so of low temperatures than those from cooler areas. |
As shown by the data below, obligatory PPs which accompany the adjectives packed, oriented, related or based can also be subsumed under so-replacement:
| a. | It was packed with cutting social observations and, even more so, with wisdom. 21 |
| b. | In general, during childhood, boys and girls are highly oriented toward their parents and less so toward their peers. |
| c. | These findings indicate that NMAE use may be related to narrative macrostructure ability and less so to narrative microstructure ability. (COCA) |
| d. | Despite their self-proclaimed commitment to human rights, aid allocations are largely based on the political objectives of donors and, less so, on the economic needs of recipients. |
For this reason, it seems viable to propose that stative -en compounds have the structure in which the left element is projected by the aP, and not by RootP. As such, they fall into the same category as compounds headed by underived adjectives such as sugar-free or oil-rich, which also feature obligatory modifiers (*free environment vs. smoke-free environment; *rich foods vs nutrient-rich foods).
|
As far as adjectives such adult-themed and heart-shaped are concerned, their derivation appears to mirror that of compounds such as long-tailed and blue-eyed, whose heads semantically refer to physical entities; hence, their syntactic representation contains nP with the adjectivizing suffix merged above.
|
The presence of nP in the structure of heart-shaped and adult-themed stems also from the fact that it is necessary for the derivation to take place. Consider the contrast below:
| a. | This box is shaped like a heart. |
| b. | This box is heart-shaped. |
| c. | This book is organized like a commentary (*by the author). |
| d. | *This book is commentary-organized. |
Both (23a) and (23c) feature stative participles, but only shaped can be considered to be derived from a noun. Hence, (23d) is ungrammatical, even though it follows the same derivational pattern.
3.3 Stative -en compounds with non-transparent meanings
Distributed Morphology posits that syntactic heads (and consequently, affixes that spell them out) may be attached directly to the root (inner affixation) or merged with an already categorized structure (outer affixation). Embick & Marantz (2008, 10) formulate the following generalizations concerning the inner and outer cycle derivation (see also Marantz 1997, 2007; Arad 2003; Embick 2010; but cf. Borer 2014).
| Cyclic generalizations: | |
| a. | Allomorphy: For Root-attached x, there may be special allomorphy, determined by properties of the Root. A head x in the Outer domain is not in a local relationship with the Root, and thus cannot have its allomorphy determined by the Root. |
| b. | Interpretation: The combination of Root-attached x and the Root might yield a special interpretation. The heads x attached in the Outer domain yield predictable interpretations. |
While compounds such as action-packed or heart-shaped are derived by the regular -ed suffix and their meanings are transparent, there exist stative -en compounds that feature irregular morphology and whose meanings are unpredictable. Examples of such compounds include spellbound, crestfallen or thunderstruck. Their syntactic behavior resembles that of compounds of the action-packed type. (25) and (26) show that they allow modification with very and suffixation with -ness: 22
| a. | I stayed outside to observe it, feeling very spellbound wanting to see what would happen next. |
| b. | (…) poor Sir Harry Towers rode away from Audley Court, looking very crestfallen and dismal. |
| c. | I looked very thunderstruck and wanted to know what was the matter. |
| a. | In any case, the students thinking can be stirred and helped to avoid one-sided spellboundedness. |
| b. | The transfer is also supposed to prompt addition of the young man's crestfallenness and sorrowful departure (…) |
| c. | Does this kind of enchantment, or smittenness, the fact of thunderstruckness, actually happen? 23 |
The stative nature of adjectives such as spellbound and their idiosyncrasy lead to the conclusion that they have a structure in which the left-most element is introduced by the Root. The attachment of the suffix directly to the root ensures a special interpretation of the structure. 24
|
At this point, the question appears whether eventive -en compounds have stative variants. While compounds such as home-made or computer-generated are occasionally found with very, they are still most likely to be interpreted as expressing results of prior events. However, very is also possible with the compounds home-made and computer-generated in contexts such as those in (28) below:
| a. | Behind the dresser is a passage which leads through to the beds. (…). It all looks very home- made and inviting. |
| b. | This font has a very computer generated feel, with quirky indents and square counters. 25 |
Neither of the compounds used in the sentences in (28) can be interpreted eventively: in (28a) home-made expresses the distinctive characteristic of the subject noun, rather than a result of an event; in (28b), the head of the noun phrase (feel) is not in a theme relationship with the verb generate. The data in (28) pose a further question of whether their syntactic representation is the same as the structure of compounds of the action-packed type. Even though a compound such as computer-generated can be used statively, it can hardly be viewed as root-derived due to the presence of the verbalizing morphology and the fact that that the left-most element has access to the event expressed by the head participle. One solution to this mismatch would be to propose that stative variants of eventive -en compounds retain all the functional structure and receive special interpretation at the head projection in environments such as those in (28), akin to how idioms are accounted for by Nanosyntax (see e.g. Caha 2009; Starke 2009; Baunaz & Lander 2018). 26
|
The problem with accounting for (28) along such lines is that the presence of the verbal structure makes it necessary for the stative variants of deverbal -en compounds to behave syntactically in a manner typical of regularly derived adjectival passives. This is not the case because, for example, home-made and computer-generated as used in (28) are impossible with agent-oriented adverbials:
| a. | *(…). It all looks veryi carefully home-madei and inviting. |
| b. | *This font has a veryi carefully computer-generatedi feel, with quirky indents and square counters. |
The data in (30) may point to the conclusion that the stative variants of eventive -en compounds have a structure without VoiceP. However, compounds which incorporate semantic agents as left-most elements can also be used as statives, as shown in (31) with man-made:
| I found these holes in a single tree, (…). They look very man-made, but could not have been. 27 |
For this reason, it seems reasonable to postulate that the stative variants of home-made and computer generated have a structure where vP and VoiceP are licensed but where there is no PASS projection, the lack of which is responsible for the impossibility of sentences in (30). Instead, the structure is headed by aP introducing the same -ed suffix used to derive compounds of the action-packed type. 28
|
The presence of the vP shell in the structure of the stative variants of deverbal -en compounds is justified by the fact that the event expressed by the verb is part of the state description, even though no such event is stipulated to have taken place. Because eventive -en compounds are necessarily derived from bi-argumental verbs, the presence of VoiceP is obligatory to ensure their transitive interpretation.
4 Conclusion
In this paper, we have argued that synthetic -en compounds in English come in two classes: eventive (which can be used either as adjectives or verbal passives) and stative (which cannot be used as verbal passives). Their contrasting features are summarized in the Table 1:
The contrasting features of eventive (resultative) and stative -en compounds
| eventive/resultative -en compounds | stative -en compounds |
|
|
The syntactic behavior of synthetic -en compounds leads to the following conclusions. Firstly, while adjectives such as student-written, pencil-drawn or home-made do project the external argument, the same is not true for compounds such as action-packed, family-oriented or heart-shaped, whose external syntax does not point to the presence of vP and VoiceP in their syntactic representation. Secondly, the non-deverbal nature of adjectives of the action-packed type entails that words equipped with passive morphology need not be eventive, even when they are compounds. Thirdly, the existence of structures such as action-packed, family-oriented or heart-shaped means that roots such as √PACK, √ORIENT or √SHAPE (as well as other roots serving as bases for the derivation of stative -en compounds) can give rise to both eventive and stative participles:
| a. | a (*carefully) heart-shaped box | (stative) |
| b. | Ancient Mesopotamia was shaped by its two rivers. | (eventive) |
Eventive -en compounds have been shown to be generally impossible as stative adjectives, with the exception of compounds such as home-made or computer-generated which can be used as eventive passives, resultative adjectives or stative adjectives:
| a. | *This map looks very pencil-drawn. | (stative) |
| b. | This map looks pencil-drawn by an artist. | (adjectival/resultative) |
| c. | This map was pencil-drawn by an artist. | (eventive) |
| a. | The furniture looks very home-made (*by an experienced carpenter). | (stative) |
| b. | The furniture looked home-made by an experienced carpenter. | (resultative/adjectival) |
| c. | This furniture was home-made by an experienced carpenter. | (eventive) |
Embick's division of participles into stative, resultative and eventive appears then to be also relevant for synthetic -en compounds. However, the agentive nature of adjectivally used deverbal -en compounds is a feature which distinguishes them from Embick's resultative participles, which are construed as not projecting the external argument.
Acknowledgements
This research was supported by grant 2018/29/N/HS2/01377 from the National Science Centre, Poland.
References
Alexiadou, Artemis , Berit Gehrke and Florian Schäfer . 2014. The argument structure of adjectival participles revisited. Lingua 149. 118–138.
Alexiadou, Artemis , Elena Anagnostopoulou and Florian Schäfer . 2015. External arguments in transitivity alternations: A layering approach. Oxford: Oxford University Press.
Alexiadou, Artemis and Terje Lohndal . 2017. The structural configurations of root categorization. In L. Bauke and A. Blümel (eds.) Labels and roots. Berlin: Mouton de Gruyter. 203–231.
Arad, Maya . 2003. Locality constraints on the interpretation of roots: The case of Hebrew denominal verbs. Natural Language and Linguistic Theory 21: 737–778.
Baunaz, Lena and Eric Lander . 2018. Nanosyntax: The basics. In L. Baunaz , L. Haegeman , K. De Clercq and E. Lander (eds.) Exploring Nanosyntax. Oxford: Oxford University Press. 3–56.
Borer, Hagit . 2013. Structuring sense, Volume III: Taking form. Oxford: Oxford University Press.
Borer, Hagit . 2014. The category of roots. In A. Alexiadou , H. Borer and F. Schäfer (eds.) The syntax of roots and the roots of syntax. Oxford: Oxford University Press. 112–148.
Bruening, Benjamin . 2014. Word-formation is syntactic: Adjectival passives in English. Natural Language and Linguistic Theory 32. 363–422.
Caha, Pavel . 2009. The Nanosyntax of case. Doctoral dissertation. University of Tromsø, Tromsø.
Davies, Mark . 2008. The Corpus of Contemporary American English. Available online at https://www.english-corpora.org/coca/.
Embick, David . 2003. Locality, listedness and morphological identity. Studia Linguistica 57. 143–169.
Embick, David. 2004. On the structure of resultative participles in English. Linguistic Inquiry 35. 355–392.
Embick, David. 2010. Localism versus globalism in morphology and phonology. Cambridge, MA: The MIT Press.
Embick, David and Alec Marantz . 2008. Architecture and blocking. Linguistic Inquiry 39. 1–53.
Embick, David and Rolf Noyer . 2007. Distributed Morphology and the syntax/morphology interface. In G. Ramchand and C. Reiss (eds.) The Oxford handbook of linguistic interfaces. Oxford: Oxford University Press. 289–324.
Halle, Morris and Alec Marantz . 1993. Distributed morphology and the pieces of inflection. In K. Hale and S. J. Keyser (eds.) The view from building 20: Essays in linguistics in honour of Sylvain Bromberger. Cambridge, MA: The MIT Press. 111–176.
Hallman, Peter . 2013. Predication and movement in passive. Lingua 125. 76–94.
Harley, Heidi . 2009. Compounding in distributed morphology. In R. Lieber and P. Štekauer (eds.) The Oxford handbook of compounding. Oxford: Oxford University Press. 129–144.
Harley, Heidi . 2014. On the identity of roots. Theoretical Linguistics 40. 225–276.
Harley, Heidi and Rolf Noyer . 1999. Distributed morphology. GLOT International 4. 3–9.
Iordăchioaia, Gianina , Artemis Alexiadou and Andreas Pairamidis . 2017. Morphosyntactic sources for nominal synthetic compounds in English and Greek. Zeitschrift für Wortbildung 1. 47–72.
Kennedy, Christopher and Louise McNally . 1999. From event structure to scale structure: Degree modification in deverbal adjectives. In T. Matthews and D. Strolovitch (eds.) Proceedings from semantics and linguistics theory IX. Ithaca, NY: CLC Publications. 163–180.
Kratzer, Angelika 1996. Severing the external argument from its verb. In J. Rooryck and L. Zaring (eds.) Phrase structure and the lexicon. Dordrecht: Kluwer Academic Publishers. 109–137.
Levin, Beth and Malka Rappaport . 1986. The formation of adjectival passives. Linguistic Inquiry 17(4). 623–661.
Marantz, Alec . 1997. No escape from syntax: Don’t try morphological analysis in the privacy of your own lexicon. University of Pennsylvania Working Papers in Linguistics 4(2). 201–225.
Marantz, Alec . 2007. Phases and words. In S.-H. Choe (ed.) Phases in the theory of grammar. Seoul: Dong-In. 191–222.
McIntyre, Andrew . 2013. Adjectival passives and adjectival participles in English. In A. Alexiadou and F. Schäfer (eds.) Non-canonical passives. Amsterdam: John Benjamins. 21–42.
Sleeman, Petra . 2011. Verbal and adjectival participles: Internal structure and position. Lingua 121(10). 1569–1587.
Starke, Michal . 2009. Nanosyntax: A short primer to a new approach to language. Nordlyd 36. 1–6.
Wasow, Thomas . 1977. Transformations and the lexicon. In P. Culicover , A. Akmajian and T. Wasow (eds.) Formal syntax. New York, NY: Academic Press. 327–360.
Throughout this paper, English synthetic compounds based on the passive participle will be referred to as ‘synthetic -en compounds.’ Additionally, the term ‘deverbal -en compounds’ will be used to refer to syntactically eventive -en compounds (that is, compounds which do project the external argument), while compounds which lack the ability to project the external argument will be labeled as ‘stative -en compounds.’
This example is quoted after Embick (2004, 357).
The fact that external arguments are incorporated into adjectival passives has been noted by Hallman (2013) who quotes examples such as NSF-funded project or voter-sponsored initiative (Hallman 2013, 83).
It should be noted synthetic -en compounds are less complex structurally than eventive participles. Specifically, synthetic -en compounds do not seem to allow the licensing of aspectual modifiers when functioning as eventive passives:
| a. | ??The picture was pencil-drawn in 10 min. (cf. The picture was drawn in 10 min.) |
| b. | ??The code was computer-generated in 10 min. (cf. The code was generated in 10 min.) |
| c. | ??The kit was home-made in two months. (cf. The kit was made in two months.) |
Example from the Corpus of Contemporary American English (COCA; Davies 2008).
Example from https://www.abs.gov.au/methodologies/provisional-mortality-statistics-methodology/jan-aug-2020.
Example from https://www.newspapers.com/newspage/98505426/.
The PASS head is adopted from Alexiadou et al. (2014); in their account, however, PASS produces both verbal and adjectival passives.
In the structure in (6), VoiceP and PASS are not cyclic heads (if they were cyclic, it would be necessary to spell out PASS with a regular suffix). Because there is no overt morphology between vP and PASS, the latter head can be spelled out with an irregular suffix, as it is the case with Tense nodes (Embick 2010).
An important implication of this is that participles such as slept-in (as used in sentences such as This bed looks slept-in) are eventive passives.
In the account of Bruening (2014), cases such as Interior still looks carved with a dull spatula (Bruening 2014, 380), whereby the passive is accompanied by an obligatory prepositional phrase, also feature an adjectival variant of the participle.
It is important to point out that not every eventive -en compound can occur in typically adjectival contexts; for example, compounds such as rain-emptied or truck-delivered are not selected by copular verbs:
| a. | *The street is/looks/seems rain-emptied. |
| b. | *The water is/looks/seems truck-delivered. |
The attributive occurrences of rain-emptied or truck-delivered appear to largely correspond to prenominal eventive passives (see Sleeman 2011).
Grammatical examples (10a–c) have been constructed by the author for the purpose of this paper. Example (10e) has been extracted from the COCA.
Note, however, that very does not always force the stative reading of a synthetic -en compound. Consider the sentence below: (i) His face was very sun bleached, his beard was grey, and what hair remained on his head had also turned grey.
Despite the presence of very, sun-bleached expresses a state having resulted from a prior event.
Example from https://en.wikipedia.org/wiki/Gay_Activists_Alliance.
Example from https://www.ryanair.com/try-somewhere-new/gb/en/travel-guides/how-to-choose-great-rome-restaurants/.
The ability of roots to take complements has generated a great deal controversy among researches within syntactic approaches to word formation. That roots can project has been claimed by Harley (2009, 2014), Borer (2013) and Iordăchioaia et al. (2017), among others. See also Alexiadou & Lohndal (2017) for an overview of the topic.
Example from https://www.nytimes.com/2020/10/13/books/review/claire-messud-kants-little-prussian-head-and-other-reasons-why-i-write.html.
Despite their overall syntactic stativity, adjectives such as spellbound, crestfallen or heartbroken are found with by-phrases:
| a. | Among those whose drug reactions drive them more toward depression and apathy, countless lives spellbound by psychiatric drugs plod along in lackluster ways. |
| b. | Crestfallen by her sister's sarcastic answer, Sarah had little time to expound as they were show into the studio. |
| c. | He was thunderstruck by what seemed like striking similarities between Sendak's childhood and his own. |
However, the agentive status of such by-phrases is questionable as adjectives such as spellbound, crestfallen or thunderstruck (in a manner similar to other -ed adjectives expressing psychological states) are also found with similarly interpreted phrases introduced by the prepositions with or at.
| a. | Brandon Hammond has left TV and movie audiences spellbound with his serious performances. |
| b. | Earlier last week, detectives were crestfallen at having to charge one of their few promising witnesses with being a fabulist. |
| c. | His family was thunderstruck with his bizarre decision and they tried their level best to bring him back to sense but to no use. |
Example from https://blog.bestamericanpoetry.com/the_best_american_poetry/2021/04/on-the-street-where-you-live-by-robert-wrigley.html.
On Distributed Morphology grounds, the present analysis cannot be extended to other OE participles because they are productively derived from their verbs and their meanings are transparent.
Nanosyntax assumes that a structure such as hold your horses receives a special interpretation at the highest node, which overrides the meanings of its individual constituents (Baunaz & Lander 2018, 33).
Adopting the structure in (32) for home-made appears to violate the generalization that cyclic heads cannot be spelled out with root-determined allomorphs. However, stative variants of eventive -en compounds are infrequent which means that one would be hard-pressed to claim that they feature regular and irregular suffixes.
