2.1. What is impersonal?

Despite having been firmly in the spotlight of linguistic studies for over a century (see van der Gaaf, Reference Van der Gaaf1904; Fischer & van der Leek, Reference Fischer and van der Leek1983; Allen, Reference Allen1986, Reference Allen1995; Denison, Reference Denison, Adamson, Law, Vincent and Wright1990; Ogura, Reference Ogura1990; López–Couso, Reference López–Couso1996; Méndez–Naya & López–Couso, 1997; Trousdale, Reference Trousdale, Fitzmaurice and Minkova2008; Loureiro–Porto, Reference Loureiro–Porto2010; Möhlig–Falke, Reference Möhlig–Falke2012; Miura, Reference Miura2015 and Castro–Chao, Reference Castro–Chao2021, among many others), there is little consensus as to what impersonal constructions are. The use of alternative labels in recent times, such as impersonal, subjectless, quasi-impersonal, nominativeless or experiencer, among others, has only served to ‘increase the already existing confusion’ (Méndez–Naya & López–Couso, 1997: 185; see also Loureiro–Porto, Reference Loureiro–Porto2010: 675–676). Moreover, although some authors talk about impersonal verbs, ‘it is more suitable to speak of personal and impersonal uses of a given verb or of verbs used personally or impersonally, rather than of personal and impersonal verbs’ (Méndez–Naya & López–Couso, 1997: 191) given that some verbs can be used in both types of constructions. Broadly speaking, there are two types of verbs capable of impersonal usage: on the one hand, verbs denoting natural phenomena, such as rain or snow; on the other, constructions with verbs referring to events which escape the volition and control of an experiencer, such as like, need or rue (see López–Couso, Reference López–Couso1996: 154 and Méndez–Naya & López–Couso, 1997: 186). It is the second group of verbs which is relevant to this study.

From a syntactic perspective, scholars tend to agree that impersonal constructions show the following two features: i) a lack of nominative argument; and ii) a verb invariably conjugated in the 3^rd person singular (see Méndez–Naya & López–Couso, 1997, and references therein). Impersonal constructions expressing non-volitional events consist of two roles or arguments: the experiencer (i.e. ‘the animate and sentient entity which perceives or experiences a concrete state’, Castro–Chao, Reference Castro–Chao, San Miguel and Neumann2018: 177), and the stimulus (also known as theme, source or cause; i.e. ‘something from which the experience emanates or by which the experience is effected’, Fischer & van der Leek, Reference Fischer and van der Leek1983: 346). Example (1) below provides an instance of an impersonal construction with a dative experiencer (him) and an accusative stimulus (hire þeawas).

1. (1) him (DAT) gelicade (SING) hire þeawas (ACC PLUR) (Fischer & van der Leek, Reference Fischer and van der Leek1983: 347)

   to him pleasure was (because of) their virtues

   (to him there was pleasure because of their virtues)

From a semantic perspective, impersonal constructions are characterised by a lack of intentionality or volition on the part of the experiencer: the experiencer, inflected for the objective case, is the passive recipient of the state or process expressed by the verb, as opposed to personal constructions, in which it is encoded in the nominative case and is the initiator of the state or process (see Fischer & van der Leek, Reference Fischer and van der Leek1983: 351 and Miura, Reference Miura2015: 10, among many others).

2.2. The demise of English impersonal constructions

Impersonal constructions were common in old Germanic languages (and also in other Indo-European languages, such as Latin, Russian and Celtic; see Möhlig–Falke, Reference Möhlig–Falke2012: 14). However, over the course of time many English verbs have changed from impersonal to personal uses. This transition seems to have been largely completed in English by the 16^th century (Fischer & van der Leek, Reference Fischer and van der Leek1983: 364 and Allen, Reference Allen1986: 401), although occasional impersonal instances are attested until about two centuries later (Möhlig–Falke, Reference Möhlig–Falke2012: 206–207 and Castro–Chao, Reference Castro–Chao, San Miguel and Neumann2018: 178). The verb to like (Old English līcian) has traditionally been used to exemplify the change from impersonal to personal (see Allen, Reference Allen1986: 375, 1995: 254; Denison, Reference Denison, Adamson, Law, Vincent and Wright1990: 113; Méndez–Naya & López–Couso, 1997: 187 and Miura, Reference Miura2015: 32).Footnote ² When trying to account for this shift towards personal use, authors have put forward a multiplicity of theories. One of the most influential accounts is that of Jespersen (Reference Jespersen1894, Reference Jespersen1927) and van der Gaaf (Reference Van der Gaaf1904): after the collapse of the English morphological system in medieval timesFootnote ³, nominative, accusative and dative were no longer distinguishable due to case syncretism. Case was therefore no longer used to mark grammatical relations. As a side effect, word order became more rigid, and the new Subject-Verb-Object arrangement contributed to the reanalysis of the pre-verbal argument as a subject. In other words, the objective experiencer became the new subjective argument, since any form in pre-verbal position is typically interpreted as a subject (see Lightfoot, Reference Lightfoot1979: 231; Ogura, Reference Ogura1990: 31 and Miura, Reference Miura2015: 9).Footnote ⁴

4.1. Reactions to the use of likey in the Web

Although myriad studies have been devoted to the verb to like and its shift from impersonal to personal (a small number of these are mentioned in Section 2 above), no work has considered the recent uses of likey. In an attempt to identify the origin of these new uses (and given that no canonical work could be consulted to this end), a first step was to turn to the Internet in order to see what language users themselves have to say about likey. It transpires that in recent years the Web has witnessed fierce debate on the use of this form and its potential racist connotations.

The etymology of the expression is not entirely clear. In his popular crossword puzzle blog, Professor Michael Sharp (aka Rex Parker) points to the potential racist roots of likey: it may have originated as a phrase of ethnic mockery of African American creole speech or, more likely, of Chinese English (Parker, Reference Parker2015). The latter theory is based on the propensity of Chinese speakers to add a long /i:/ <ee> to many English words. Exaggerated representations of Chinese speakers with this phonological feature were common in fiction and the mass media of the 19^th century, where Chinese speakers ‘were ridiculed ad absurdum’ (Mieder, Reference Mieder1996: 7). The main target of such a slur were Chinese laundrymen, as working in laundries was common for this ethnic group of immigrants (according to Mieder, Reference Mieder1996: 11, by 1920 about 30% of the Chinese in work in the United States carried out duties related to laundry work). It is precisely in this context where the proverb ‘no tickee, no washee’ (which may have had a decisive influence on the emergence of ‘me [no] likey’) originated in the 19^th century. In his play Two Men of Sandy Bar (1876), Hop Sing depicts an

Even though the proverb ‘no tickee, no washee’ seems to derive from the attempt to mock the Pidgin English of Chinese immigrants, not all speakers are aware of the racist history behind it. In all likelihood, the formula underwent semantic generalisation (Pfenninger, Reference Pfenninger2009: 14–15) and took on a wider figurative meaning, which allowed its use in new contexts to indicate that a certain prerequisite was required in order to obtain a given result (see Partridge, Reference Partridge1977: 157; Wilkinson, Reference Wilkinson1993: 378 and Mieder, Reference Mieder1996: 14). Accordingly, the formula may have been reanalysed as a partially fixed expression of the type ‘no X, no Y’ with no racist undertones, as in ‘no pain, no gain’ or ‘no ID, no entry’, implying a conditional relationship between the lack of X and the lack of Y. After the generalisation of meaning, ‘no tickee, no washee’ may have influenced the emergence of a new, somehow related expression, namely ‘X likey’, where the long final /i:/ is also present. As the historical data in Section 4.2 below will show, the spelling variant likee used to combine with the negative particle no more often than with any other linguistic element (especially in the 19^th and early 20^th centuries), probably replicating the negative structure of ‘no tickee, no washee’.

However, other theories do not contemplate racism as the origin of likey. In his blog on linguistics, Professor Arnold Zwicky (Reference Zwicky2012) discusses the expression ‘me no likie’ and adds two more possibilities to the list which have nothing to do with ethnic mockery: it may just be an expression which echoes childish language (that is, the way a three-year-old child would speak) or may even have originated in the animated TV comedy Family Guy, where the sequence is used in several occasions (in one of the scenes, Lois Griffin shouts ‘me likey bouncy, me likey bouncy!’ while jumping on a trampoline; in another scene, she says ‘me likey breadsticks, me likey . . . ’; Sheridan & Hogan Reference Sheridan and Hogan2001). This latter theory is the one also proposed in Urban Dictionary (2003) under the entry for ‘me no likey’. How can we reconcile these four theories on the origin of likey? There may be a certain degree of plausibility to all of them, as will be shown in the following section.

4.2. Likey in COHA

In this section, we turn to COHA in order to find historical evidence on the use of likey. First, the morpho-syntactic features of the sequences where likey is used are described, and then some examples are considered in greater detail so as to check the theories on its etymology described above.

All the examples in COHA (45) date from 1846 to 1934. As regards spelling, the corpus search returned two examples of likey and 43 of likee, thus showing that final <-ee> was the prevailing spelling in this database. In most of these examples, the verb takes a pronominal subject (21 examples, 47% of the total; see [2] below), but its use with no explicit subject (14 examples, 31% of the total; see [3]) is also rather common, followed by nominal subjects (10 examples, 22%; see [4]). Figure 1 compares the use of pronouns in the different persons.

1. (2) Me likee Americanos. (COHA, 1912, Rolling Stones, by O. Henry)

2. (3) Me cook-man in Melican army. No likee war. (COHA, 1921, Grace Harlowe's Overland Riders on the Great American Desert, by Jessie Graham Flower)

3. (4) Mellican man no likee Chinaman hab 2 wifee. (COHA, 1885, Ah Sin, by Mark Twain)

Figure 1. Pronominal forms used with likey in COHAFootnote ⁵

As Figure 1 shows, likey exclusively takes singular pronouns in the historical data. It is also more frequently used with first-person pronouns, especially in the objective case. In fact, in 11 instances the objective pronoun is me (see [2] above and [5] below), and only in two examples it is a different form, in this case him (see [6]):

1. (5) But the white man was convulsed with fear, and said nothing in the making ready of the boat, not even ‘No, no’ when Salesa put her arms round him and kissed him again and again on the lips; and Billy Hindoo shook like a wet dog in the bow, whimpering, ‘Hi, yi! me British subject! me no likey!’. (COHA, 1921, Wild Justice: Stories of the South Seas, by Lloyd Osbourne)

2. (6) I heap’ shamed. You fightee my China boy, you catchee me. My boy no mo’ hab me fo’ boss -savvy? I go back, him no likee me. Mebbe all same killee me. (COHA, 1898, Moran of the Lady Letty, by Frank Norris)

As for the type of sentence in which likey appears, it is more common in negative sentences (28 examples, 62% of the total), followed by declarative sentences (16 examples, 36%) and questions (just one token, 2%). Moreover, the only negative particle used in COHA is no rather than not. All this may be related to the conditional use of the formula ‘no tickee, no washee’ seen in Section 4.1: the earliest attestations of likey may be more common in negative sentences (always with ‘no’ as a negative particle and never accompanied by an auxiliary form) because of an influence of the formula ‘no tickee, no washee’, also a negative construction. Finally, likey is more common in shorter sequences, especially in those consisting of between two and four words (29 examples, 64% of the total).

Moving on to a more qualitative reading of the examples, we can see that likey is repeatedly used in excerpts which show a situation of contact between white and either Afro-American (see [5] above, where Billy Hindoo is described and referred to as ‘a nigger’) or especially Chinese (see [3], [4] and [6]) characters, which is in keeping with the idea of likey originating from the slur on racial stereotypes seen in Section 4.1 above. There is a third instance which deserves closer attention. Example (7) is an excerpt taken from Otto Jespersen's well known 1928 work on language. Here he connects two of the aforementioned theories on the origin of likey, namely childish language and the caricaturisation of Chinese speakers of English. From his words we can derive a certain degree of disdain towards these two kinds of languages (i.e. childish and Pidgin) as, for him, they are two impoverished or corrupted linguistic forms:

1. (7) But many parents are not so wise; they will say Btinged themselves when once they have heard the child say so. And nurses and others have even developed a kind of artificial nursery language winch they imagine makes matters easier for the little ones, but which is in many respects due to erroneous ideas of how children ought to talk rather than to real observation of the way children do talk. [ . . . ] I give a connected specimen of this nursery language (from Egerton, // Keynotes, 85): ‘[ . . . ] Hitchy cum, bitchy cum, bitchy cum hi, Chinaman no likey me’ This reminds one of pidgin-English, and in a later chapter we shall see that that and similar bastard languages are partly due to the same mistaken notion that it is necessary to corrupt one's language to be easily understood by children and inferior races. (COHA, 1928, Language: Its Nature, Development and Origin, by Otto Jespersen)

In short, historical evidence from the corpus material shows that likey is attested since as early as the 19^th century. Even though it is not possible to give an overview of the kind of narrative in which this construction proliferates due to the low number of authors using it, it becomes patent that travelogues and novels concerning colonial activity feature this construction heavily. Given that in these examples it was mostly (if not always) used by non-native (especially Chinese but also Afro-American) characters, the origin of the expression seems in fact to be marked by ethnic connotations. Based on the corpus evidence, we can conclude that expressions with likey did not originate in the series Family Guy, although their use in the series may have put them into hyperdrive in recent times.