Pozna ń It ’ s Raining Immigrants ! HELLelujah ! : The Metaphors of Immigration in Early American Magazines ( 1828 – 1959 )

Stemming from a conviction that the same phenomenon can be construed diff erently by diff erent cognisers, metaphors used “refl ect[ing] and eff ect[ing] underlying construal operations which are ideological in nature” (Hart 2011, 2), the present paper investigates how the conceptualisation and linguistic construction of changed over time, forwarding a convenient representation of reality. To that end, the study marries the Cognitive Linguistic approach to Critical Discourse Analysis (Charteris-Black 2004; Hart 2010; 2011; 2015) with the multifactorial usage-feature analysis (Glynn 2010). The results have shown that in the times of increased migration were objectifi ed, their otherness foregrounded through appropriate discursive strategies and topoi. Curbing immigration in later periods contributed to an observable shift in the linguistic representation of the immigrant out-group.


Introduction
The outbreak of the Arab spring, 1 as well as the so-called Arab winter (Mihaylov 2017) into which it evolved in a number of the Arab League countries, resulted in the largest wave of immigration since WWII (Kingsley 2015). Risking their lives more often than not, migrants tried to cross the Mediterranean to seek asylum on the shores of Malta and Italy. The unprecedented scale of the population displacement was a matter of concern to many, a fact that did not go unnoticed by rightwing populists across Europe. Not infrequently, conservative, nationalist forces attempted to score points by stoking up the fear of "the other", a strategy which proved eff ective as a springboard to political power in more than one EU member state. As a basic emotion (Ekman 1999), however, fear knows no geographical or administrative boundaries.
Playing the same (race) card may have, therefore, aff ected voting decisions overseas in 2016, when Donald Trump was elected the 45 th president of the United States. In his campaign, described by The New York Times as "explosive, populist and polarizing" (Flegenheimer and Barbaro 2016), Trump did not hesitate to use infl ammatory rhetoric reproducing harmful, negative immigrant stereotypes. Discursively constructing immigrants as trash, animals and rapists (Korte and Gomez 2018), he triggered in his base specifi c conceptualisations which helped him impose the desired representation of reality. Surprising as the deployment of such discourse might have been in a country built, after all, by immigrants, Trump's discriminatory remarks only replicated claims American history had known before.
O'Brien's paper on the language of the early 20 th -century immigration restriction debate in the U.S. provides ample evidence of the dehumanisation therein of the so-called "new immigration" 2 through the use of , , and metaphors. The present study completes O'Brien's by supplementing the diachronic and quantitative dimensions. Stemming from a conviction that the same phenomenon can be construed diff erently by diff erent cognisers, the metaphors used "refl ect [ing] and eff ect [ing] underlying construal operations which are ideological in nature" (Hart 2011, 2), it investigates how the conceptualisation and linguistic construction of changed over time, forwarding a convenient representation of reality. To that end, the present contribution marries the Cognitive Linguistic approach to CDA (Charteris-Black 2004;Hart 2010;2011) with the multifactorial usage-feature analysis (Glynn 2010). While the former draws on the tools of Critical Discourse Analysis, Cognitive Linguistics, corpus linguistics and pragmatics to uncover the intentions of language users, the latter helps operationalise the shades of meaning of the lexeme which profi les the target domain in question, contributing to the description of its conceptual structure.
The structure of the remainder of the paper, in turn, is as follows. Section two provides an overview of literature on the linguistic representation of immigration in historical English. Section three discusses the approach adopted for the purpose of the study. Sections four and fi ve contain the analysis proper, preceded by the description of the data set and the annotation procedure. The fi nal section of the paper off ers concluding remarks together with suggestions for further research.

The discursive representation of immigration in early English: a literature review
Unsurprisingly, publications dealing with the representation of immigration in historical English are few and far between. For one thing, its scale and fi erce intensity, which is what may have elevated the status of immigration to a research question in the fi rst place, is a product of relatively recent times. For another, the pool of Present-Day English data readily available is so vast and the problem so pressing that a steady stream of scholarly eff orts is naturally channeled into revealing the intentions of the privileged actors controlling the narrative. Among the notable exceptions are papers by O'Brien (2003), Park (2006;2008) andmost recently -Bremmer, Jr. (2019). O'Brien's "Indigestible Food, Conquering Hordes and Waste Materials" (2003) deals with the metaphorical themes structuring the early 20 th -century immigration restriction debate in the United States which culminated in the implementation of the Immigration Acts of 1921 and 1924. It was, apparently, a marked increase in the numbers between 1880 and 1920, the clustering of newcomers in large urban areas and a shift in their provenience that added considerably to the immigration-related fears and anxieties. Throughout the debate new immigrants were dehumanised through a systematic use of the , , and metaphors. Framing them as possibly poisonous, alien food, "cargoes of human fl otsam" or "human refuse", a body of water, swarming bees, mosquitoes, ants and parasites (O'Brien 2003, 36-43) paved the way for the introduction of more restrictive legislation. To convince the public of its validity and advisability, the new immigrant class had to be constructed as standing in stark opposition, if not posing a threat, to whatever American citizens may have held dear.
A similarly polarised, dichotomous representation of immigration is what Park (2006) discusses in detail in her paper on the discursive construction of immigrants in US social work between 1882 and 1952. The study examines how the public views of a select, infl uential elite among social workers, voiced in three US social work periodicals popular at the time, helped forge and reproduce immigrant identities in and through a discourse of hierarchical opposites (Park 2006, 173-174). In line with Derrida's approach to meaning as "established through diff erentiation" (Park 2006, 172), it was through binaries 3 that immigrants were linguistically constructed against the background of the (superior/sophisticated/ industrious) native stock. As the Author observes, throughout the period in question the American-ness could only be delineated and defi ned in reference to that which was dismissed as alien. Interestingly, which side of the binary exactly a given group fell on at a given time changed, depending on the broadly understood socio-political context. Consequently, "the incoming Celts were reviled, for example, until they were outnumbered by the infl ux of Mediterraneans and the Hebrews; these Eastern and Southern Europeans were found dangerously inferior, except when measured against the irredeemably defective Asiatics" (Park 2006, 175). Inclusion was, therefore, best not taken for granted, the privileging of one people or class always paralleled by the denigration of the other.
That same shifting valorisation and its responsiveness to the dynamics of the socio-political reality surfaces two years later in yet another paper by Park, namely "Making Refugees: A Historical Discourse Analysis of the Construction of the 'Refugee' in US Social Work, 1900-1957. Devoted largely to the issue of the problematisation of refugees in the wake of WWII, it falls, however, outside the scope of interest of the present contribution. For even though the Author does brief the reader on the mechanisms of framing immigrants in the disciplinary discourse, they only serve as the other pole opposite the concept of "refugee" (Park 2008, 773).
Binary oppositions as a discursive device for constructing otherness also feature in Bremmer, Jr.'s "'Thi Wilde Witsing': Vikings and Otherness in the Old Frisian Laws" (2019). Admittedly, the paper does not deal with the linguistic representation of immigration, but it does share with Park's studies an elaborate discussion of how the identity of the Other (here: an invader, a Norseman) is constructed in discourse in contrast to the familiar, the good, (and the southern) (Bremmer, Jr. 2019, 16). In the Twentieth of The Twenty-four Landlaws, a set of 13 th -century regulations specifying the legal relations among the Frisians, the Vikings are represented as "the ultimate 'Other'", dangerous, brutal and ruthless (Bremmer, Jr. 2019, 14). And as they physically penetrate and thus dishonor the integrity of the Frisian territory; […] utterly shame a man by depriving him of his right as a free man to go wherever he wants to […] bind him against his will, abduct and hold him captive; […] force their victim to join them on their raids and commit the three greatest crimes that Frisian land knows, called haveddeda (capital crimes): homicide, rape and arson -and not just burning down any old houses, but even God's houses… (Bremmer, Jr. 2019, 24-25), their identity thus defi ned presupposes the existence of its opposite, a valiant, south-oriented 4 Frisian.
The present paper has been designed as an addendum to O'Brien's, therefore -like his -it focuses on the metaphoricity of the immigrant discourse at the turn of the 19 th century. Rather than selectively sample illustrative evidence in support of one's argument, however, it narrates the development of the conceptualisation of immigrants on the basis of annotated corpus data with a view to answering the research question, i.e. how the concept evolved between 1828 and 1959 as manifested in American English.

The theoretical model
Given that a considerable part of our conceptual system is metaphor-based (Steen 2014, 117), researching the (chronological) evolution of a concept involves, of necessity, studying cognitive metaphors. Originally, i.e. back in the 1980s, cognitive metaphors were understood as thinking about one (typically more poorly delineated) domain in terms of another, more concrete and better delineated, e.g., or , as in They've been through a lot, the collapse of the Chinese economy. However, decades of research resulted in a more detailed model of metaphor. Conceptual metaphors of the kind mentioned above have come to be viewed as complex cognitive structures built from neurally entrenched primary metaphors such as, e.g., or . While experientially motivated primary metaphors are still believed to consist in a cross-domain mapping from the source (more concrete) concept to the target domain, complex cognitive metaphors may be processed diff erently (Steen 2014, 126; for a discussion see Steen 2014). In this paper, metaphor is taken to be a primarily discursive phenomenon, both refl ecting and infl uencing social structures (Hart 2015, 322).
Whether understood as a cross-domain mapping, a class-inclusion statement, embodied simulation or a blend, metaphor is a multi-aspectual mechanism 5 (Cameron 2007) powerful in its ability to arouse emotions and aff ect the way we think, experience and act (Lakoff and Johnson 1993, 4). If that be true, their aff ective force and evaluative nature should make metaphors effi cient and eff ective carriers of ideology. This is particularly true of conventional metaphors, the long-established, well-entrenched ways of talking and thinking about phenomena in the world. Their danger and potency lies in the fact that they work in the background, triggering specifi c conceptualisations largely unbeknownst to the recipient. Consequently, "the 'logic' in the target domain is not consciously experienced as derived and therefore mediated but is taken for granted as absolutely, objectively refl ecting reality" (Hart 2015, 325). Mental representations thus prompted, and accepted as own, naturally constitute addressees' experience of the issues under discussion (Hart 2011).
Notwithstanding their (infamous) role in constructing ideology, however, metaphors can inform Critical Discourse Analysis, i.e. a study of cognitively mediated relations between discourse and society (Van Dijk 2009, 64). The same metaphorical linguistic expressions which carry ideological implications provide access to the conceptual metaphors involved in the activation of particular "evaluatively loaded frames or scripts […] against which the profi led concept is understood" (Hart 2010, 8). Cognitive processes, inaccessible directly, lend themselves to deconstruction through careful analysis of linguistic structure, whose choice is key, as Hart (2010, 2-3) observes after Langacker, in that alternative structures encode competing conceptualisations.
The Cognitive Linguistic approach to CDA has long had its advocates, notably Charteris-Black (2004), Hart (2010;2011;, and Musolff (2012). While Charteris-Black (2004, 8) underscores the signifi cance of cognitive semantics in providing uniform criteria for classifi cation and comparison of metaphors across discourse domains, Hart (2010;2011; highlights the potential of Cognitive Linguistics to disclose ideologically-charged construals evoked by the use of meticulously tailored discursive strategies. In a similar vein, Musolff (2012, 301) speaks of the role cognitive metaphor analysis plays in detecting and exposing racist ideology. At the same time, permeating the four texts is the omnipresent belief that, to be more socially relevant, any analysis of metaphor should take into account its pragmatic role and persuasive function (Charteris-Black 2004, 8-9;Musolff 2012, 301). In other words, it is essential to examine metaphor not only as a cognitive tool, but also as one of pragma-linguistic devices (Molek-Kozakowska 2014) instrumental in the realisation of discursive strategies. Since attributing "particular ideas and values to a target domain" (Maalej 2007, 136) is among the tasks performed by metaphor in discourse, the present study, if cognitive in principle, also looks at the predicational strategies and the topoi employed.
Commonly used when writing or talking about the Other those "reservoirs of generalized key ideas from which specifi c statements or arguments can be generated" (Richardson 2004, 230) feature prominently "in argumentation for and against discrimination" (Reisigl and Wodak 2001, 75). Reisigl and Wodak (2001, 75-80), Wodak (2009, 44) or Richardson and Wodak (2009, 49) list the most typical content-related argument schemes found in migration discourse. Here belong, for example, the topoi of danger or numbers, which stress the necessity of adopting preventive measures in the face of potential dangers or threats, especially if confi rmed by the numbers. The topos of authority, along with the topos of law, justifi es actions performed based on the premise that whatever "the authority" decides is right. The topos of culture, in turn, can be explicated in the following way: "because the culture of a specifi c group of people is as it is, specifi c problems arise in specifi c situations" (Reisigl and Wodak 2001, 79-80).
Predication, which -next to reference and argumentation -serves as a means of positive self-and negative Other-representation, assigns a specifi c prejudiced trait or characteristic to the target out-group as representative of it (Reisigl and Wodak 2001, 55). Among the most frequent discriminatory predications realised linguistically in texts about immigrants are the claims that they stand out beyond integration (strategy of dissimilation), that there are too many of them (aggregation), that they are inclined to criminal activity (criminalisation) or that they are dirty and unlearned (primitivisation), to name but a few. Reisigl and Wodak (2001, 55) observe that such "prejudices are used in argumentation, functionally employed as premises -strictly speaking, as fi rst parts of the conclusion rules".
To properly complement O'Brien's, this paper combines Critical Metaphor Analysis with what Glynn (2010, 8-9) describes as a multifactorial approach to language or multifactorial modeling. The attractiveness of the approach lies in its tools which enable the researcher to measure the seemingly unmeasurable. What it means is that multivariate statistics provides the means to operationalise, and thereby capture, the interaction between the many layers or aspects of language use, including meaning. Linguistic phenomena indiscernible via introspection undergo statistical analysis with the aid of specially designed techniques which identify and translate the co-occurrences in the data into patterns of usage, themselves further translatable into patterns of conceptual structure (Glynn 2010). Admirably, they can also help verify the statistical signifi cance of the results as well as test the explanatory power of a given analysis (2010,17). What is absolutely wonderful about the method is that, apart from a range of formal criteria such as aspect, tense or number, it is possible to incorporate and quantify the socio-cultural dimension of language. This requires a careful, manual annotation of the sample, a procedure which is painstaking, time consuming and -of necessity -characterised by a considerable degree of subjectivity. Once annotated, however, the data can be treated with quantitative techniques producing objective, replicable and falsifiable results. There is also a bonus: the analyses examine both the infl uence of the factors upon language use as well as the mutual impact those factors exert upon one another.

Data and the coding schema
The Corpus of Historical American English (henceforth COHA) is a structured 400 million-word database, the largest of its kind, balanced by genre decade by decade. The texts come from the 1810s-2000s and fall into one of four categories: fi ction, magazine, newspaper and non-fi ction books (Davies 2010). Between 1828 and 1959, in the "magazine" section of COHA the noun immigrants occurs 750 times. On the grounds of constituting "linguistic noise" 14 instances had to be removed upon closer reading. Each of the remaining 736 occurrences has been carefully analysed in context. Qualitative analysis in this case meant annotating every instance for the century, predicational strategy, speaker evaluation, topos and the source domain. Table 1, below, shows the feature labels used in annotation and the categories found:  *19c2 = 1826-1850, 19c3 = 1851-1875, 19c4 = 1876-1900, 20c1 = 1901-1925, 20c2 = 1926-1950, 20c3 = 1951-1959 The coding schema as presented in Table 1 includes all the categories as attested in the data. Yet, for statistical analysis only those categories which were represented with the frequency of 8 or higher could be considered. These are marked in bold. As a result, only 626 sentences were used as data input for correspondence analysis (see Table 2). As can be seen in Table 2 immigration was most often discussed between 1875 and 1950, that is in the period of intense immigration from China and Eastern Europe. Most of the sentences show relatively neutral speaker evaluation and about 65% of them do not contain any metaphorical expressions. Among those that do, the most common source domains are represented by , , and . All sentences retrieved from the corpus have been found to realise a predicational strategy, with as many as 90% employing an argumentation strategy, i.e. a topos. Examples below illustrate selected values of coding categories: (1) We fear nothing from the political or sectarian machinations of these immigrants; and did no other danger threaten us from this source… (North American Review, 19c3) (speaker evaluation: negative, : non-metaphorical, predicational strategy: dissimilation, topos: danger) (2) Irish, Germans, Swedes, and Norwegians, no matter how poor, have always been welcome. They once formed the bulk of our immigrants. (South Atlantic Quarterly, 20c1) (speaker evaluation: positive, : , predicational strategy: ethnifi cation, topos: numbers) Sentence (1) uses the topos of danger as expressed through such lexemes as fear and threaten. The use of these lexemes contributes to the construction of the negative speaker evaluation. It is further strengthened by the dissimilation strategy as expressed with the NP sectarian machinations and the use of the demonstrative these. Excerpt (2) has been coded as positive, based on the adjective welcome conveying the impression of desirability. The speaker objectifi es the immigrants ( ) by the use of the phrase the bulk of which also contributes to the topos of numbers. The reference to nationalities: Irish, German, Swedes and Norwegians exemplifi es the predicational strategy of ethnifi cation.

Analysis and discussion
The paragraphs below discuss how each of the categories, i.e. the source domain, predicational strategy and topos, contributes to the construction of the immigrants in public discourse.
The source domains identifi ed in the data set by and large serve construing immigrants as inanimate beings ( / ), coming in large numbers, devoid of personality and therefore dispensable and easily replaceable, as shown in examples (3)  The as a , blatantly referred to as cargo in example (6), were not only imported, but also supplied, e.g., (8) That leaves us, then, to consider the lands of Eastern and Southeastern Europe, which, with 11111y, tillye supplied the majority of immigrants to America since 1890. (Saturday Evening Post, 20c2) ( : ) (9) Or should we go on limiting European immigrants to a maximum of about 150,000 a year, while welcoming any and all of the peoples of the three Americas, the latter having supplied about one fourth of the immigrants who have come here since 1929? (Saturday Evening Post, 20c2) ( : ), and as is very often the case with products, either purchased or received as gifts, their quality may leave a lot to be desired: (10) The pauper, the lewd woman, the criminal, the diseased person, th' person who had sold his labor, we could complain of as unsatisfactory immigrants.
(North American Review, 19c4) ( : ) The metaphor, manifesting in the use of the verb receive (examples 11-12), by no means testifi es to the desirability of the transferred "object" or the satisfaction of the recipient: (11) We are not now organized to receive immigrants properly, or to distribute them in such a way as to insure ourselves against aggravation of the present evils of increasing unemployment and labor unrest. (New Republic, 20c2) ( : ) (12) The American continents can no longer receive immigrants in great numbers.
(Saturday Evening Post, 20c2) ( : ) Rather, it furthers the depersonifi cation of immigrants, together with the metaphor motivating examples (13) and (14): (13) Unless Europe makes every eff ort to improve its own position the overseas countries will not be in the best position to absorb immigrants. (Harpers, 20c2) (14) Now faced with the problem of absorbing 100,000 new immigrants (mostly from Communist Rumania), the government last week slapped new taxes, up to 70%, on consumer goods, ranging from aspirin to refrigerators. (Time,20c3) Nowhere in the data set is the substance identifi ed, but it is easily interpretable as a liquid, especially when viewed against the country construed as a (15): (15) They emphasize the progress Zionism has brought to Palestine, especially in the form of economic improvement shared by Jews and Arabs alike; they demand that Jewish immigrants be accepted to the full absorptive capacity of the country. (Nation, 20c2) Yet, examples (16) and (17) from Atlantic suggest a diff erent, non-liquid reading: (16) But against these advantages must be set the diffi culty of dealing with masses of immigrants, wholly untrained… (Atlantic, 19c4) ( : ) (17) So far as individual race traits are concerned, it would seem that there is no especial trouble to he apprehended from the mass of our newest immigrants.
(Atlantic, 20c1) ( : ) Precisely what the substance is, however, matters less than what its cognitive job is, i.e. to take the focus away from the individual, to obscure motives and intentions (Hart 2010) and thereby discourage empathy (Charteris-Black 2006). The metaphor, in turn, realising the predicational strategy of naturalisation, evokes the topos of danger, covertly communicating the need to curb the fl ow so as to regain control, e.g., (18) The great waves of immigrants who preceded the Puerto Ricans came here in clusters and settled in national colonies where a man could live a lifetime without learning English. (Saturday Evening Post, 20c3) (19) No doubt this charge is not entirely unfounded, but hitherto the constantly swelling stream of immigrants has supplied most of the laborers for the rougher kinds of work, and the young men educated here have found plenty of room higher up the economic ladder. (Atlantic, 19c4) (20) With the stupendous demand for unskilled labor of every description, women were pressed into service, and still the supply came nowhere near meeting the demand, and vast fl oods of immigrants began to pour in upon our shores.
(Good House, 20c1) (21) Only an emergency makeshift of the Department of State in interpreting strictly the law excluding "persons likely to become a public charge" is now thinning the tide of immigrants that could easily be more than 150,000 a year from Europe. , ) which would seem to highlight the agency aspect of the target domain by structuring as animate organisms. This agency is construed as ability to move as is the case with swarming bees or -at a higher level -the ability to make decisions and follow them ( , ): (22) They are the sons and daughters, the grandsons and the granddaughters of the immigrants who swarmed into Boston in the last quarter of the nineteenth century. (Saturday Evening Post, 20c2) ( : ) (23) It was utterly alien, however, to many people of colonial stock, and to the great armies of immigrants who swarmed to our hospitable shores. (Atlantic, 20c2) ( : ) (24) But the race from which the fi rst immigrants spring has also a powerful, and perhaps more enduring infl uence… (North American Review, 19c3) ( : ) Construed as or (examples 22-24), immigrants, perhaps, come across as less passive or powerless, but the suggestion of evil lurking in the shadows is present none the less. Only the metaphor, and there are three such instances in the material, takes the stigma of undesirability off of immigrants. Still, a guest -no matter how welcome or useful -is never at home and his stay has a "best before" date: (25) It is individual capitalists who have worked the mines, made the railroads, invited the immigrants and lent them money to improve their farms. (Atlantic,19c4) ( : ) Understandably, most of the predicational strategies employed (notably dissimilation, ethnifi cation, racialisation, primitivisation and religionisation) highlight otherness of the newcomers to the point of "not belonging", e.g., (26) If there are immigrants radically alien as a race, socially and morally, to American civilization, their case must be decided by the same paramount rule of the public good, care being taken that the interest of the state is not confounded with industrial rivalry or inhuman antipathy of race. (Atlantic, 19c4) (ps: dissimilation) (27) Our liberal immigration laws have been severely criticised. The immigrants are illiterate, poor, badly nourished and unskilled. It has been said that "they are beaten men from beaten races, representing the worst failures in the struggle for existence," degraded, criminal, quiescent, lacking initiative and responsibility, or the capacity for taking advantage of new opportunities; these immigrants have a low standard of living, which they are not inclined to improve if they prosper; and they are likely in a few years to fi ll our almshouses, insane asylums, and hospitals. (South Atlantic Qarterly, 20c1) (ps: primitivisation) The ones that remain (aggregation, depersonifi cation, and naturalisation, in particular) objectify immigrants and stoke anxieties, especially when coupled with a particular set of recurring topoi (see Reisigl and Wodak 2001, 75-80), e.g., It is hardly surprising that of the topoi identifi ed in the data, the vast majority are either those tapping into societal fears of heightened crime rate (topos of crime), spread of disease (topos of danger), abuse of the welfare system (topos of burden/fi nances), irreconcilable diff erences (topos of culture/character) or being outnumbered (topos of numbers), e.g., The emergent discursive representation of immigrants as the Other is therefore far from idyllic, with a strong inclination towards negative stereotypisation. Even though the topoi of usefulness (36-37) and success (38-39) are likely to trigger a more positive response in text readers, and the strategy of victimisation (40-41) certainly can evoke sympathy, overall pre-1960 American magazines foreground a biased view of immigration: (36) …but the gain in this respect has been small in comparison with what the immigrants were worth as laborers in the varied branches of industry.
(Scribners, 19c4) (37) In two respects the absorption of large numbers of immigrants from many nations into the American commonwealth has been of great service to mankind. (Atlantic, 19c4) (38) Immigrants, already established in America, sent home money to permit other immigrants to come. (Harpers, 20c1) (39) In spite of certain talk we have heard lately from the unthinking, America is proud of its immigrants. (Time, 20c2) (40) The Communists constantly harass the immigrants, said Bishop Ladyka, and try to bribe them to return to Europe. (Time, 20c2) (41) The laborers were all immigrants, mostly Mexicans and Italians. If a laborer was injured, he lost his job. (Atlantic,20c2) In order to answer the question how the conceptualisation of changed over time, the annotated data have been treated with Multiple Correspondence Analysis (MCA). MCA is an exploratory technique that "reveals frequency-based associations in corpus data […] and visualises these associations to facilitate their identifi cation […] The visualisation of the relations takes the form of confi guration biplots, or maps, which depict degrees of correlation and variation through the relative proximity of data points (which represent linguistic usage features and / or the actual examples of use" (Glynn 2014, 443-485). Figure 1, below, illustrates the correlations in the data set signifi cantly contributing to the model. The right-hand bottom quadrant of the plot shows how the use of the metaphor correlates with the use of the topos of danger and the strategy of naturalisation. At the same time the positioning of those data points relative to the other variables suggests their being distinctly dissociated from the other features. This is unexpected, but possibly attributable to the fact that constructing immigration as dangerous, moving water is the timeless, fl agship property of discourse on the displacement of populations (Santa Ana 2002;Charteris-Black 2006). The top right-hand quadrant displays an accumulation of linguistic strategies instrumental in reproducing negative Other representation. If those were to be translated into the then immigration-related concerns they would be the ethnic/racial and cultural background of the newcomers and their fi nancial status. The source of those anxieties were the wave of immigration from Ireland and Germany between 1820 and 1860 as well as the infl ux of Chinese workers, which began in 1850s, and the arrival of over 20 million "new" immigrants from Central, Eastern and Southern Europe between 1880 and 1920. A look at the bottom left-hand quadrant, in turn, shows that the introduction of the legislation restricting immigration to the United States (the Chinese Exclusion Act of 1882 and the Immigration Acts of 1891, 1917, 1921 and 1924), may have helped ease the xenophobic fears. 20c2 and 20c3 correlate with largely non-metaphorical uses, natural speaker evaluation and such topoi that point to amicable coexistence in place of hostility and reluctance.