Metaphoric representations of the migration crisis in Hungarian online newspapers: A first approximation*

1. Introduction

The study of political metaphors has always been a popular field of interest within metaphor research (cf. Schön 1993, Mio 1997, Santa Anna 1999, Charteris-Black 2006, Musolff 2015, 2016, Brugman et al. 2017), thus it does not strike one as surprising that the migration crisis of 2014-2015 has also launched various research projects on the topic (e.g. Spieß 2017). Several papers and smaller articles addressing the communication surrounding the migration crisis and its visual and linguistic representations have also appeared in Hungary (e.g. Bernáth & Messing 2015, Fülöp et al. 2017). However, these studies merely list the metaphors appearing in Hungarian public discourse as illustrative examples; so far, no methodologically firmly founded, comprehensive, and systematic data gathering has taken place, which would provide a data set for the purposes of further metaphor analyses which would also be relevant from the perspective of metaphor research.

Accordingly, based on linguistic data (on a corpus of app. 35 million words compiled by the authors) the present paper ventures to (i) identify the exact time period of the migration crisis that is relevant for metaphor research, (ii) to present a representative amount of corpus data from this period (iii) with the help of a novel method developed by the authors (iv) for the identification of the metaphors defining Hungarian public discourse on migration, to serve as input for further qualitative, quantitative, and comparative analyses both across languages and across various media, genres, and registers.

The theoretical framework of our investigations is a strongly revised version of Conceptual Metaphor Theory as put forward by Schwarz-Friesel (2013, 2015). Furthermore, our study relies on the analysis of political metaphors offered by, among others, Musolff (2004), Charteris-Black (2006), Schwarz-Friesel & Kromminga (2014), Schwarz-Friesel & Reinharz (2013), and Ziem (2008). For the purposes of the present paper, we consider political metaphor as a cognitive heuristic tool which helps a given linguistic-cultural community to access, understand and communicate complex and intricate social and political situations that would otherwise be difficult for individuals to interpret based on their own everyday experiences and knowledge (cf. Mio 1997, Charteris-Black 2006).

The relevant data are extracted by a semi-automatic procedure, the so-called “funnel” method developed by one of the authors (Majoros 2013, 2016). The method has been inspired by Jäkel’s onomasiological approach (1997, 2003) and integrates elements from other corpus-based methods of metaphor research (cf. Stefanowitsch 2006). The recursive application of the method allows for the exploration of the central metaphors we are looking for, whereas we can compare and contrast the metaphorical language use of each sub-corpus based on frequency data.

2. Data and procedure

2.1. Corpus

The corpus we have compiled consists of the articles published between 1 July 2014 and 31 December 2015 on two major Hungarian news portals, to be more precise, in the online version of two serious mainstream print newspapers (Heti Világgazdaság, henceforth: HVG[1], a socially-politically oriented national weekly and Magyar Nemzet, a renowned national daily, henceforth: MNO[2]) without users’ comments. For gathering the texts, we used a so-called crawler or spider application designed specifically for this purpose.[3] Each (sub-)corpus constituted by the texts of the two online news portals can be examined both separately and together, as well as in a six-month and a monthly distribution; furthermore, the complete text of each article can be retrieved, if necessary. This enabled us to examine the entire context of relevant expressions and phrases and to track the potential changes in metaphoric source domains over time during the 18 months under scrutiny. For the management and analysis of the corpus (collocation analysis, concordancing, compiling KWIC-lists, etc.) we used the freeware corpus analysis toolkit AntConc, developed by Laurence Anthony.[4]

            In the corpus linguistic literature, text corpora of between 1 and 100 million words generally qualify as representative (cf. Stefanowitsch 2005: 114). The corpus we have compiled contains in total of 36,062,001 words (tokens); thus regarding its size it can be considered statistically representative. The articles comprising the corpus are concerned with various topics of public interest (such as domestic and international news, economy, culture, science and arts, technology, sports, etc.).

2.2 Procedure

Apart from some technical specifics, during the analysis of our data we relied on the idea and process behind the so-called “funnel” method (cf. Majoros 2013), although we had to consider the technical possibilities offered by AntConc.[5] The “funnel” method is a semi-automated procedure designed for the comprehensive examination of metaphors in a particular field. Its most useful achievement is that data gathering is not based on an introspectively compiled list of metaphoric expressions or on dictionary entries but the lexemes needed for the formulation of the search queries come from the corpus itself. Beyond the quantitative revision of earlier qualitative analyses, the procedure is capable of a systematic survey of metaphors used in any area of experience. By applying the procedure, the introspective nature of the research methodology can also be decreased or minimized. The procedure can be considered a “hybrid” technique, integrating several aspects of various, already known procedures (for a more detailed overview of these see Stefanowitsch 2006, or more recently Semino 2017). The application of the “funnel” method can be sketched out in the following steps.

2.2.1 Step 1: Identification of the relevant context excerpts 

As a first step, we can immediately start out with an automatic search, if we define a search query including one or more linguistic expressions that are somehow related to the target domain under scrutiny (TD items). By applying step one, we can not only narrow down the scope of our corpus to the texts and passages relevant for the target domain but are also able to examine the immediate context of the chosen TD items. We assume the contexts of the TD items to constitute a sequence of texts that are potential hubs for conceptual metaphors characteristic of the target domain in the form of linguistically realized metaphors with source domain (SD) items as their vehicles. However, unlike earlier corpus linguistic procedures in metaphor research (cf. Stefanowitsch 2006), the aim of step one in the “funnel” method is not merely to find TD items that are embedded in metaphoric linguistic expressions. Although the processing of the data starts with a search for one (or more) TD item(s), the outcome of the initial search is not merely a list of the TD items’ collocations but it is supplemented by a corpus of the relevant passages that is based on a list of collocations and is substantially smaller than the entire original corpus. This corpus of manageable size consists of passages (i.e. broader context excerpts, in our present case, of whole articles) containing the TD items.[6]

            In the present study, the first step was to conduct a search for the search expression migráns* as a TD item in both sub-corpora (HVG and MNO) broken down into 18 months. Since the form migráns* stands for the string of characters m+i+g+r+á+n+s followed by any number of any characters in the search syntax of AntConc, we were able to find all forms of the word migráns (inflected, derived, or in compounds). Public discourse in Hungarian revolving around the topic of migration uses in general two other nouns referring to the people arriving in Europe: bevándorló (‘immigrant) and menekült (‘refugee’). These three nouns (migráns, bevándorló, and menekült) are far from being referentially equivalent or being near synonyms regarding their connotations – on the contrary, the framing of the whole issue may depend on which one is chosen. However, we had to exclude bevándorló and menekült from our investigation for practical reasons stemming from their grammatical-morphological features: the word bevándorló is homonymous as a noun and as the present participle of the verb bevándorol (‘to immigrate’), whereas menekült is homonymous as a noun and as the part participle of the verb menekül (‘to flee, run away, take refuge’). Their inclusion would have meant numerous irrelevant hits that would have had to be excluded manually. Furthermore, in this way we were able to avoid the possible distortion of our results due to the differing connotations and underlying concepts behind the three nouns.[7] The results of the first step in the data gathering process determined the following steps of the whole procedure. They are summarized in Figure 2.1.

Figure 2.1 – The frequency of migráns* in HVG/MNO between July 2014 and December 2015

The diagram shows the search results yielded by the search expression migráns* in the whole corpus broken down into 18 months between 1 July 2014 and 31 December 2015.[8] As can be seen, the results suddenly sky-rocket in the second half of 2015 in the case of both online newspapers, and their distribution over the 18 months under scrutiny is roughly identical in both sub-corpora.

            In the light of the frequency data in both online newspapers, it can be safely stated that in 2014 and in the first months of 2015 the migration and the arrival of migrants were still topics of negligible importance for public discussion in Hungary. The numbers first stand out in June 2015 both in HVG (265 hits) and in MNO (175 hits). Since this month – as the parallel increase of the frequency data on both online portals attest – the topic has gradually but very swiftly come to the centre of attention. In August 2015, the frequency abruptly doubles in the case of both newspapers. The absolute high is apparently reached in September 2015, when the July numbers nearly quadruple. The last quarter of 2015 shows a gradual decrease in relation to September, at a similar pace and in parallel in both sub-corpora. An interesting difference between the two sub-corpora is that the October values are behind those of August in the case of HVG, while in the case of MNO they substantially exceed them. Furthermore, in the case of HVG the month of November shows a sudden, significant decrease in the frequency of migráns* compared to October, while this change is not so sharp in the MNO sub-corpus. However, it must be noted that over the last three months of 2015 the number of hits in the MNO sub-corpus exceed those in the HVG sub-corpus significantly, even though its absolute size is less than that of HVG in the same period. The difference between the two sub-corpora becomes most apparent in November, when the number of hits in the MNO sub-corpus is more than twice the number of hits in the HVG sub-corpus.

            Based on the apparent findings presented in Figure 2.1, we focussed our attention on the period between June and December 2015 in the case of both sub-corpora in order to uncover the metaphors used in the conceptualization of the migration crisis. Since the expression migráns* occurs in September 2015 in both sub-corpora with an excessively high frequency, we conducted a collocational analysis for all forms of the word migráns in the articles published in this month. We reviewed the collocational list, identified all collocations that may qualify as metaphoric, and exported manually all the contexts (articles) in which they occurred. In this way, when searching for possible SD items we did not have to rely on our intuitions and introspective analysis, on dictionaries or on previous research, since the collocational list and the contexts belonging to the possibly figurative collocations yielded a smaller corpus (i.e. a series of relevant contexts/articles) which later served as the source of the SD items we were originally looking for.  

2.2.2 Step 2: Identification of the relevant SD items

Step two involves the manual reading and checking of the exported context excerpts in order to find and collect the metaphoric linguistic expressions occurring in them. In other words, the aim of step two is to identify the SD items that are used metaphorically (in our case: expressions that in fact appear as an SD item of a metaphor whose target concept is the migration crisis/the migrants). During this phase, the linguist does not necessarily read and analyse coherent, entire texts (the whole original corpus) but only those excerpts of the contexts (passages, articles) that are relevant for his search. This significantly decreases the amount of work that must be done manually and speeds up the processing of the data. The drawback of the procedure so far is that obviously not every single metaphoric expression can be filtered out from the corpus; however, a list of SD items can be compiled that can be considered representative of the TD item (i.e. the particular target concept of the metaphoric conceptualization, modelling, or framing under investigation) and can serve as a starting point for the next step of the procedure.

            In the present study, this meant that we manually examined all exported contexts, identified the relevant SD items, agreed upon their metaphoricity, and set up a list of search expressions that included all possible forms of the SD items we had found (Table 2.1).

Table 2.1 – Search expressions from the manually identified SD items to be applied in Step 3[9]

2.2.3 Step 3: Extending the search for SD items

In the third step, we extend the search for the SD lexemes identified in the previous step to the entire corpus (or to other corpora if applicable, as for instance in the case of a comparative corpus analysis). In this phase of our investigation, we do not need to formulate any guesses or initial assumptions regarding the possible SDs exploited for the conceptualization of the TD since the search conducted in the entire corpus involves only those context-relevant SD items that we have already identified. In other words, the search in step 3 is based exclusively on the linguistic data gathered at the end of step 2 (search expressions) and on the twelve most frequent forms of the noun migrant in our corpus (context words, see Table 2.2).

Table 2.2 – The twelve most important forms of the lexeme migráns in the corpus

Ideally, when following the lines of the “funnel” method all forms of migráns could have been used as a context word but our opportunities were limited by the functionality offered by AntConc. In AntConc, we could not import a list of all word forms of migráns and use them as context words; to do this we would have had to type in all inflected, derived, and compounded forms one by one in the case of each and every search in the case of every month. Since this would have required considerable time and effort, for this reason we decided to use only the twelve most frequent forms in the corpus. The context words include five compounds (migránsválság, ‘migrant crisis’; migránsválsággal, ‘with the migrant crisis’; migránskérdés, ‘migrant question’; migránsügyben, ‘in the migrant issue’ and migránsáradat, ‘flood of migrants’), the latter of which was obviously in itself metaphoric.

             The contexts where one of the search words (SD items) and one of the context words (a form of migráns) co-occurred were automatically retrieved from the whole corpus. The scope of the query was eight words both left and right from the context word, i.e. we searched for passages where one of the search expressions occurred within a distance of eight words from one of the context words. The KWIC lists yielded by the automatic search for each month was than checked and reviewed manually and we applied labels for the tentative SDs we encountered.[10] A further aim of the manual review of the KWIC lists generated for each month was to find so far possibly unidentified, new figurative elements (i.e. those not covered by the search expressions) and draw conclusions about other novel SDs based on these SD items. Nevertheless, the manual review of the co-occurring search items was indispensable since it prepared step four, i.e. the conversion of the results into numeric data.

2.2.4 Step 4 and beyond: Classification of the SD items and conversion into numeric data

After classifying the relevant SD items into SDs and labelling them accordingly (flood, war, object, pressure/burden, animal, building, disease, explosion, and miscellaneous), we counted the frequency of all metaphoric expressions and those of each SD broken down into the 18 months under scrutiny so that we could get a picture of how dominantly each metaphor appears in the corpus.[11] Based on these frequency data, we could draw conclusions as to what metaphors play a key role in the conceptualization of the migration crisis and migrants. These conclusions may serve as a starting point to formulate hypotheses regarding the metaphoric structure of the texts or the thematic corpus under scrutiny.

            However, it must be noted that data processing cannot be equated with the interpretation of the data, thus a final, fifth step is needed, which points beyond the relatively theory-independent procedure offered by the “funnel” method. This closing step cannot be considered to be a part of the latter any more, since it cannot be conducted without relying on a specific model of what metaphor is. Although the procedure we applied ends in an interpretative phase, we will not go beyond the description of our findings and as a theoretical consideration we must simply record that the data we present below can be best grasped, analyzed and interpreted along the theoretical assumptions and notions put forward within the framework of the cognitivist paradigm of metaphor research (cf. Kertész et al. 2012).

            In sum, the alternative procedure outlined in this section functions like a funnel since during its application the linguistic material under scrutiny is gradually narrowed down to the point where we actually work with numbers instead of specific linguistic expressions. The initial search in step 1 was conducted on the entire corpus, while step 2 involved only those passages (articles) that were considered as relevant. Step 3 concentrates exclusively on the SD items found, and then in a final step we present and evaluate the frequency data. The procedure is “hybrid” in the sense that automated phases are followed by manual analyses.

3. Findings

The procedure described above yielded a total of 588 metaphoric expressions in our whole corpus. The metaphoric expressions identified are distributed over the period under scrutiny (see Figure 3.1 below) in accordance with the frequency data of the wordforms and compounds of the noun migráns (migráns*) (see Figure 2.1 above). The first month where we have encountered metaphoric formulations regarding migration and/or migrants was February 2015 (5 relevant tokens). However, the number of metaphoric occurrences remains fairly insignificant all through March (3), April (3), and May (7). The summer months show a slight and steady increase (June – 18, July – 29, August – 50), while in September the number of hits increased exponentially. October, November, and December show a gradual decrease but the migration crisis and migrants seem to remain a central topic of Hungarian public discourse as metaphorized target concepts.

Figure 3.1 – The number of metaphoric expressions found between February and December 2015

Based on the monthly distribution of the frequency data yielded by the search expression migráns* and that of the metaphoric expressions identified in the corpus, it can be safely stated that the topic of the migration crisis exploded into the focus of Hungarian public discourse in September 2015 and remained at the centre of attention all through the rest of the year. This seems to suggest that the period especially relevant for metaphor research is the second half of 2015.

            The metaphoric expressions we have identified seem to be arranged around seven major source domains: flood (200 tokens), war (116 tokens), disease (89 tokens), object (84 tokens), pressure/burden (42 tokens), animal (34 tokens), and building (18 tokens) (see Figure 3.2).

Figure 3.2 – The number of metaphoric expressions in each source domain

Apart from these major source domains for the conceptualization of the migration crisis/migrants we have found two expressions where the start of the crisis appears as an explosion (a migránsválság kitörése ‘the outbreak of the migrant crisis’):

(1) A migráns-, illetve menekültválság kitörésekor ugyanakkor keresztények széles rétegei jól érzékelhetően nem tudtak mit kezdeni azzal, amikor a bevándorlók segítői a képükbe vágták a Máté-evangélium idézetét: „idegen voltam, és befogadtatok”.[12]

(2) De nem az ilyen retorikai túlzásoknak köszönhette mostani választási győzelmét, ahogy a menekültügynek se (a „migránsválság” kitörésekor már rég nagyon vezetett a közvélemény-kutatásokban).[13]

Since the explosion-metaphor seems to represent a rather marginal and isolated case, these will be left out of consideration.

            The remaining three metaphoric expressions instantiate a parasite-metaphor as well as historical analogies:

(3) A német döntéshozók a jelek szerint tisztában vannak azzal, hogy nagy tévedés a migránsokat élősködő tömegnek tekinteni.[14]

According to the signs, German decision makers are aware that it is a grave mistake to consider migrants as a parasitic crowd.

(4) Angela Merkel német kancellár azt szeretné, hogy Lengyelországban „a migránsok számára koncentrációs táborokat alapítsanak”.[15]

The German Chancellor, Angela Merkel wants “concentration camps to be set up for the migrants” in Poland.

(5) […] Németország "a bevándorlás földje" – migránsokat nagy tömegben befogadó ország […][16]

[…] Germany is “the Promised Land of immigrants” – a country admitting migrants on a large scale […]

The parasite-metaphor has been reported by several authors to be recurrent in political and public discourse on immigration (most notably Musolff 2015 and 2016: Chapter 6), however, it is almost completely absent in our data set. This apparently surprising fact may be explained on the one hand by the nature of our corpus, i.e. the online version of a moderate and reasonable daily and weekly newspaper. On the other hand, the parasite-metaphor presupposes at least some temporary co-existence of the immigrants and the native population of the admitting country. This first-hand experience is virtually non-existent in Hungarian society since there are no significant immigrant populations in the country. The historical-cultural analogies appearing in (4) and (5) would require a detailed analysis of their own but since their weight seems to be negligible in comparison with the other major source domains in the present paper we must put them aside.

            Our discussion of the concepts and domains used to talk and think about migration and migrants in the Hungarian press will exclude the source domain of disease. In 88 cases out of 89, these expressions involve the verb kezel (‘treat, handle, manage’). Kezel is a highly polysemous verb across several domains from medicine and machines to people and problems with highly conventionalized and lexicalized meanings, thus it is rather a dead metaphor which has nothing to do with conceptualizing migrants as pathogens or the migration crisis as a disease. The only expression where the migrants might be understood as the reason behind a physical symptom in an otherwise healthy organism would be (6) referring to railway traffic stoppages:

(6) A magyar vasutat megbénító, harmadik világbeli migránsok […]

Migrants from the third world, paralyzing the Hungarian railway […]

            In the following, we discuss in more detail the remaining six major source domains in the order of their weight in our data set as illustrated in Figure 3.3.

Figure 3.3 – The frequency of metaphoric source domains in %

3.1 Flood-metaphors

Beyond doubt, the most central and most dominant source domain in the metaphoric conceptualization of the migration crisis is the domain of flood, i.e. migrants appear as an uncontrolled and uncontrollable mass of water. Slightly more than a third of all identified metaphoric expressions (200 tokens) instantiate this metaphor. It is not only the frequency data which speak for the postulation of the flood-metaphor as the Leitmetapher of Hungarian press articles on the migration crisis, but also its distribution across the 11 months under scrutiny (Figure 3.4).

Figure 3.4 – The frequency of the flood-metaphor (February-December)

Its first occurrence is documented as early as March. There are only two months where it is absent and from August on it seems to be the image used most often to talk about migration.         Despite its dominance, the flood-metaphor seems to be exploited rather poorly in the sense that the mapping of the source domain onto the target is not very elaborate and involves only a few structural elements and aspects of the source domain alongside some emotive connotations. Accordingly, the linguistic manifestations of the flood-metaphor are rather homogeneous in the form of various synonymous nouns and their N+N compounds designating massive bodies of water (hullám ‘wave’, áradat ‘tide, torrent’, özön and özönvíz[17] ‘deluge, flood, high water’, folyam ‘stream, river’) and verbs referring to the intensive movement of water (sodor ‘to carry away, wash away’, eláraszt ‘to flood, overflow’, (be)áramlik ‘to stream, flow (in)’, (be)özönlik ‘to stream, flow, flood (in)’, (rá)zúdít/zúdul ‘to shower, pour, flow (onto)’, feltorlódik ‘to dam up’).[18] Apart from these metaphoric expressions, we encounter only a minority of phrases fitting the metaphoric model flood, such as the following:

(7) […] azonban egyelőre nincs bizonyíték arra, hogy a szélsőségesek valóban beszivárognának a migránsok közé.[19]

[…] however, for the moment there is no evidence yet that extremists would indeed infiltrate into the migrants.

(8) […] a Frontex munkatársai […] az illegális határátlépők kiszűrésében és a migránshullám kezelésében segítenek a migránsok regisztrálásával.[20]

[…] the employees of Frontex […] assist in filtering those illegally crossing the borders and in handling the wave of the migrants by registering them.

(9) […] az intézkedés célja, hogy a migránsokat az áteresztőpontokra, a legális határátkelőhelyekre tereljék […][21]

[…] the purpose of the action is to drive the migrants to the checkpoints, to the legal border crossing points […][22]

(10) […] nem lehetne meggátolni a térségünkbe telepített migránsokat, hogy Nyugatra menjenek feketézni […][23]

[…] it would not be possible to keep the migrants resettled into our region from working illegally in West-European countries […][24]

These less frequent source domain items do not elaborate on the metaphor further but simply manifest the metaphor the migrants are a massive body of water / the migrant crisis is a flood.

Considering its emotional connotations and its inferential potential, the flood-metaphor highlights and conveys among others the following aspects of the target domain: (i) the large number of migrants, (ii) their arrival is uncontrolled and uncontrollable, (iii) it imposes a grave danger on us (just like a natural disaster) and that (iv) it must be stopped (somehow). However, these attitudes and metaphoric inferences are not expressed explicitly in the form of more heterogeneous linguistic metaphors, suggesting a more specific and detailed elaboration of the conceptualization behind the metaphor on a linguistic level.

3.2 War-metaphors 

Regarding its frequency, the second most dominant metaphor defining public discourse about the migration crisis is the war-metaphor (116 tokens, 20% of all metaphoric expressions found). Its distribution over time is similar to that of the flood-metaphor except for a more abrupt decline in number following September (see Figure 3.5).

Figure 3.5 – The frequency of the war-metaphor (February-December)

            However, the variation of linguistically manifest source domain items seems to be larger than in the case of the flood-metaphor. In the war-like framing of the migration crisis, migrants appear as aggressive invaders arriving in great numbers, laying siege to the borders with the intent of conquering our countries, and against whose attacks we must protect ourselves:

(11) „Ezek nem menekültek, ez egy invázió” […][25]

“These are not refugees, this is an invasion”

(12) […] a magyar embereket és Magyarország területét megvédje az illegális migránsok tömegeitől […][26]

[…] to protect Hungarian people and Hungary’s territory against the crowds of illegal migrants. […]

(13) „A migránsok megszállták a déli határainkat […]”[27]

“The migrants have occupied our Southern borders […]”

(14) […] a kancellár Németország határainak a megnyitásával migránsok rohamát indította el […][28]

[…] by opening the borders of Germany, the Chancellor has started an attack/storm of migrants […]

(15) „[…] mi ezt a háborút elveszítettük”.[29]

“[…] we have lost this war”.

(16) […] a menekültek utánpótlása kimeríthetetlen.[30]

[…] the reserves of the refugees are inexhaustible.

            As this small sample taken to illustrate the war-metaphor suggests, the emotional effects and inferential potential of the metaphor is far more intensive than that of the flood-metaphor. It centres the attention of the public on (i) migrants as a large hostile army (ii) arriving with the intent of aggression and invasion, who (iii) threaten our very existence, and against whom (iv) the last lines of defence are our borders (v) which must be protected by any means. The war-metaphor seems to trigger a strong “king of my castle” attitude and it communicates that our castle is under siege. However, many of the expressions found instantiate conventional, lexicalized meanings, such as határvédelem or a határok megvédése ‘border protection/management/security’ which are neutral rather than combative or aggressively warlike and the most intensive and militant passages we have found are quotes from and references to views expressed by politicians and other prominent public figures. In other words, despite the dominance of war-metaphors, press accounts seem to be rather moderate and considerate in their use.

3.3 Objectification

Our category set up for object-metaphors subsumes figurative expressions where migrants are handled as inanimate objects that can be freely manipulated. After September 2015, these metaphoric expressions seem to remain an integral part of the discussions on the handling of the migration crisis (see Figure 3.6).

Figure 3.6 – The frequency of object-metaphors (February-December)

Despite our expectations, apart from a few highly creative and innovative metaphors such as (17), we have found that migrants are not conceptualized as various specific objects but merely appear as the object of certain verbs referring to their management as inanimate entities without a volition of their own as in (18) or (19).

            (17) A nagy európai mingránspingpong[31]

            The great European migrant ping-pong

(18) A visegrádi négyek nem hisznek abban, hogy a krumpliszsákként ide-oda tologatott migránsok az EU nagy tervéhez alkalmazkodni fognak – olvasható a cikkben.[32]

The Visegrad Group does not believe that the migrants pushed back and forth as a sack of potatoes would conform to the grand plan of the EU – says the article.

(19) Így a célországok szépen kiszemezgethetnék a képzett migránsokat […][33]

Thus, the target countries could pick out the highly qualified migrants as it pleased them […]

Apart from these rare examples, the data we have found appear almost exclusively in the context of the European quota plans aimed at dividing refugees between the member states, based on some predefined criteria. They all involve a form of the verb oszt (eloszt ‘to divide, spread’, szétoszt ‘to distribute, divide up’, újraeloszt ‘to redistribute’) or its synonyms (szétszór ‘to spread, disseminate’) with the noun migrant(s) as their object.

These object-metaphors are somewhat incongruent with those previously discussed: the uncontrollable and aggressive migrant masses appear here as physical objects that can be manipulated easily. Yet, it must be kept in mind that the flood- and war-metaphor seem to represent a perspective closer to the stance of Hungary in the debate on handling the influx of migrants which emphasizes the possible dangers of the situation, whereas the object-metaphor apparently represents the more optimistic stance of the EU regarding the issue. 

3.4 Pressure/burden-metaphors

The pressure/burden-metaphor is one of the first metaphors to appear in our corpus (3 tokens as early as February 2015). It is also present in all months except for March, although with a lesser frequency and intensity than the flood- or war-metaphor. Based on Figure 3.7, we can state that the frequency of the pressure/burden-metaphor is relatively evenly distributed around the absolute high of September over the six remaining months (June-August and October-December) in the second half of 2015.

Figure 3.7 – The frequency of the pressure/burden-metaphor (February-December)

Unlike objectification, the pressure/burden-metaphor seems to be well compatible and coherent with the other metaphors found. It can most notably be linked to the flood-metaphor (the image of a massive body of water exerting pressure on the dams), but it can also be related to the object-metaphor (since most physical burdens are objects too):

(20) […] az EU-ban már menedékjogot kapott migránsokat is szétosztaná az egyenlő teherviselés elve alapján.[34]

[…] would also divide up the migrants who are granted asylum in the EU based on the principle of equal burden sharing. 

In example (20), the above-mentioned conceptualization of migrants as inanimate objects is clearly connected with handling the migration crisis/migrants as coming to grips with a physical burden. Migrants are often conceptualized as a weight; they (over)burden the administration, the asylum system, and the budget, and they are seen as a work load.

            In the following passage, we encounter an even more specific image pointing beyond the concept of mere physical burden as a source domain. Here, it is connected to the pressure exerted by the influx of migrants evoking a flood-control scenario.

(21) Tiszteletben kell tartani Magyarországot és döntéseit a nagy nyomás miatt, amely a migránsok beáramlása miatt nehezedik rá […][35]

We must respect Hungary and the decisions it has made because of the pressure weighing down on it due the inflow of migrants […]

This metaphoric model is often specified further and applied to particular situations as attested by the following illustrative example:

(22) […] csökkent a nyomás a horvát határon, […][36]

[…] the pressure has decreased at the Croatian border, […]

Here, the Croatian border also appears as a section of a flood-protection dam system under imminent water-pressure within the flood-control scenario. This kind of integration of flood- and pressure-metaphors into a single image is in general characteristic of the metaphoric passages found in the corpus.

3.5 Zoomorphic metaphors

The distribution of animal-metaphors over time seems to be a bit different from that of all the other source domains: on the one hand, its distribution is less equal, and on the other, its absolute high is not in September but in August. Figure 3.8 shows a more rhapsodic picture than that of the source domains discussed earlier.

Figure 3.8 – The frequency of animal-metaphors (February-December)

However, it must be noted that similarly to the disease-metaphor mentioned at the onset of the presentation of our findings, where the verb kezel was almost exclusively the sole basis of metaphoricity, the linguistic expressions labelled as belonging to zoomorphic metaphors are centered around the forms of the verb terel ‘to drive, shepherd, herd or flock into a direction (together, in, out)’ with various verbal prefixes (el-, össze-, kiterel) and the compound terelőkordon ‘a cordon marking the way of passage’.

This means that no animal is named explicitly that would serve as a source concept for the metaphoric conceptualization of migrants, but the choice of the verb terel suggests that the groups of migrants are perceived as herds or flocks of domestic animals. Although the primarily lexicalized meaning of terel in Hungarian is to watch over and drive animals into a particular direction as they are grazing, it is also used with objects designating masses of bodies or groups of animate objects, thus all hits with the verb terel may also be linked to the flood- and the object-metaphor (e.g. in the sense of diverting the movement of a mass or crowd). In sum, the linguistic manifestation of the animal-metaphor is poorly elaborated but it is often combined with other metaphors:

(23) A kerítésépítés nem megoldás, de olyan szükséges és elkerülhetetlen lépés, amely megpróbálja törvényes és ellenőrzött keretek közé terelni az illegális migránsok áradatát […][37]

Building the fence is not a solution but it is a necessary and inevitable step that tries to herd the flood of illegal migrants into a legal and controlled framework […] 

(24) […] a kerítés nem fogja sem érdemben megállítani, sem elterelni a migránsok hadát […][38]

[…] the fence will not be able to effectively stop or divert (literally: drive off/ away) the army of the migrants […]

As can be seen, the of-possessive constructions realizing the flood- and the war-metaphor (the flood and the army of the migrants) respectively are linked to the animal-metaphor as the object of the verb terel and elterel.

3.6 Building-metaphors

Despite the small amount of the data, it can be clearly seen in Figure 3.9 that the building-metaphor constantly decreases in the months following the absolute high in September.

Figure 3.9 – The frequency of the building-metaphor (February-December)

Despite its relatively low frequency, the building-metaphor appears in a more varied fashion in the corpus. We encounter a classic image where Hungary (25) and the European Union (26) appear as a building. Many typical aspects of this metaphoric model seem to be exploited in the context of the migration crisis: we encounter opened and closed doors, gates, windows, and corridors:

(25) A migránsok nem eleve nyitott ajtón jöttek be az országba.[39]

The migrants did not enter the country through a door that had been open for them from the beginning.

(26) Miután Dél-Európa egyre inkább bezárja kapuit, a migránsok új útvonalon próbálnak eljutni Észak-Európába.[40]

Since Southern Europe is starting to close off its gates, the migrants are trying to get to Northern Europe on a different path.

In a similar vein as the metaphors discussed above, the building-metaphor also seems to be combined or even “contaminated” by other metaphoric models:

(27) […] amikor az újkori limest hiába ostromló migránsáradat inkább egy könnyebb úton hömpölygött tovább.[41]

[…] as the tide of the migrants laying siege to the modern limes in vain streamed along an easier way.  

(28) Ezeket már rég ki kellett volna dolgozni, jóval azelőtt, hogy a migránsok áradata az ajtónkon kopogtatna.[42]

These should have been developed long, long before the flood of the migrants came knocking on our door.

In (27), we find a strange “mega-blend” of the metaphors flood, war, and building, whereas in (28), the combination of the flood- and the building-metaphor results directly in a mixed metaphor or a catachresis.

4. Conclusion and possibilities for future research

The most important result of our study is the successful implementation of the “funnel” method in the identification of the metaphors of the migration crisis in Hungary in 2014-2015. With the help of the procedure sketched in Section 2 we have uncovered the most prominent metaphors in the conceptualization of the migration crisis in the Hungarian online press: the flood-, the war-, the object-, the press/burden-, the animal-, the building-metaphor.

If we intend to describe these prominent metaphors we found in the corpus, then we can consult Chilton’s (1996) distinction between major and minor metaphors, or Santa Ana’s (1999: 198) description of major metaphors as a class of metaphors whose (i) source domain is similar, (ii) which show a relatively high frequency and (iii) have a great variety of linguistic realizations. The first two criteria for classifying the above-mentioned metaphors as major figures shaping the public mindset in Hungary regarding the migration crisis are obviously met: we have managed to delineate tokens of metaphors with a similar source domain occurring with a high frequency, yet the variation of their linguistically expressed form is strongly limited, since they tend to be centered around a couple of SD items.

Moreover, these SD lexemes seem to be highly lexicalized and polysemous across domains, e.g. the verb véd across the domains of flood-protection, border-management and war, or the verb terel across the domains of grazing animals and massive bodies of water. Journalists, news writers, and column writers seem to pick up only a relatively small portion of the possible linguistic SD items, they do not elaborate, expand, or creatively vary the metaphors found. They tend to pick out a few catch phrases from a domain such as flood, war, or animals, and use them in their metaphors all over again. This may be associated with the property of press texts that they are less diverse regarding their figures of speech (cf. Santa Ana 1999: 198); they probably make use of stronger images put forward by politicians and other influential opinion leaders and re-use them. The result of this is that the metaphoric expressions we have uncovered are not very elaborate, creative, or intensive, in the sense that they do not seem to convey expressly offensive, derogatory, and denigrating meanings at first sight. This finding is in accordance with Musolff’s (2015 and 2016: Chapter 6) comparative analysis of the use of parasite-metaphors in the mainstream press, in online fora, and in blogs, which suggests that the former seems to represent the most moderate form of public discourse about socio-political issues.

Nevertheless, the metaphoric patterns we have found almost exclusively convey a negative attitude towards migrants. They suggest that migrants are looked upon as a grave danger and a threat (in the flood- and the war-model) and as an undifferentiated mass, not as a group of individuals (in the object-, the animal-, the flood-, and the war-model). All metaphors uncovered imply a strong us-against-them perspective: if they are the flood, we are the ones in danger; if they are the invaders, we are the ones who must protect ourselves; if they are objects or animals, we are the ones who can manipulate them; and if they are animals, they must be somehow below us as human beings. These models all reflect our fears. Empathy towards the migrants seem to be missing: the fears that they are trying to escape from and the uncertainty awaiting them in Europe remain unexpressed.

The moderate use of metaphoric language by the press seems to indicate that the inferential, persuasive, and emotive potential of single metaphoric expressions is relatively low, i.e. the tokens we have found are not detailed or elaborate in and of themselves but, as can be seen, their mere conventionality and frequency might trigger a series of metaphoric inferences. They might not be intensive separately, in the sense that they are not explicitly negative and offensive but their effectiveness might reside in their totality, their high frequency and homogeneity. As Santa Ana points (1999: 203) out with reference to Chilton and Ilyn (1993), if political metaphors show some rivalry and they vary, they will be less fixed parts of public thinking, whereas when public discourse is determined by uncontested, frequent, and exclusively used metaphors the metaphoric conclusions they imply will become a productive way to conceptualize the issue at hand.

With the help of our method, we have also been able to reinforce all metaphoric models that have been discussed earlier in the literature on metaphor and migration. Even a brief look at only some of the previous results makes it apparent that metaphors used in entirely different historical, geographical, cultural and socio-political contexts resurface in the Hungarian public discourse about migration. Santa Ana (1999) examines 107 articles published in the Los Angeles Times between June 1992 and December 1994 and identifies almost without exception those metaphors that we have found: animal, burden, disease, dangerous water, and war. In O’Brien’s (2003) study of US public discourse about immigration restriction around the turn of the 19^th and 20^th century, we encounter among others the same fears and metaphors as those occurring in our Hungarian corpus: disease, object, natural catastrophe (most notably: flood), as well as animal and sub-human metaphors.[43] Dervinytė (2009) conducts a comparative analysis of English and Lithuanian press texts published between 2004 and 2008 and arrives at the same conclusion we do: the two most prominent metaphors in public discourse about migration are water and war. Various researchers might use different labels, or their analyses might differ in their specifics but the very same cognitive mechanisms and linguistic expressions seem to be at work here.

It is an important achievement of our paper to take an initial step towards the systematic, meticulous, and empirically based linguistic identification of metaphors in public discourse about migration in Hungary. It presents a structured data set by applying careful methodology for further research without sticking rigidly to any theoretical framework. One of the many possibilities for future investigations might be the question of how the distribution of our data over time relate to real world political events and to particular milestones in government communication – statements, reports, or interviews – about the migration crisis. Another fruitful line of investigation might be concerned with the functions and effects political metaphors exert on individuals and on public opinion (cf. Boeynaems et al. 2017). The data presented in this paper might serve as useful input for comparative studies in any cognitive based framework across genres and languages, both in a qualitative and a quantitative vein.

References

Anthony, Laurence (2016): AntConc (Version 3.5.0) [Computer Software]. Tokyo, Japan: Waseda University.

Bernáth, Gábor / Messing, Vera (2015): Bedarálva. A menekültekkel kapcsolatos kormányzati kampány és a tőle független megszólalás terepei. In: Médiakutató 16-4, 7–17. [Available at: http://www.matarka.hu/klikk.php?cikkmutat=2520798&mutat=http://epa.oszk.hu/03000/03056/00061/pdf/EPA03056_mediakutato_2015_t%C3%A9l_007-017.pdf (30.10.2017)

Boeynaems, Amber / Burgers, Christian / Konijn, Elly A. / Steen, Gerard J. (2017): The effects of metaphorical framing on political persuasion: A systematic literature review In: Metaphor and Symbol 32-2, 118–134.

Brugman, Britta C. / Burgers, Christian / Steen, Gerard J. (2017): Recategorizing political frames: A systematic review of metaphorical framing in experiments on political communication. In: Annals of the International Communication Association 41-2, 181–197. Available at: http://www.tandfonline.com/doi/full/10.1080/23808985.2017.1312481 (30.10.2017)   

Charteris-Black, Jonathan (2006): Britain as a container: Immigration metaphors in the 2005 election campaign. In: Discourse and Society 17-6, 563–582.

Chilton, Paul (1996): Security Metaphors: Cold War Discourse from Containment to Common House. New York: Peter Lang.

Chilton, Paul / Ilyin, Mikhail (1993): Metaphor in political discourse: The case of the “Common European House”. In: Discourse and Society 4-1, 7–31.

Dervinytė, Inga (2009): Conceptual emigration and immigration metaphors in the language of the press: A contrastive analysis. In: Kalbų studijos / Studies about Languages 14, 49–55. Available at: https://www.kalbos.lt/zurnalai/14_numeris/09.pdf (30.10.2017)  

Fülöp, Nóra / Ilyés, Virág / Katona, Eszter / Morvay, Gergely / Varjú, Zoltán (2017): A migráció arcai. In: Nyelv és Tudomány. Available at: https://www.nyest.hu/facesofmigration/ (30.10.2017)

Jäkel, Olaf (1997): Metaphern in abstrakten Diskurs-Domänen. Eine kognitiv-linguistische Untersuchung anhand der Bereiche Geistestätigkeit, Wirtschaft und Wissenschaft. Frankfurt am M.: Peter Lang.

Jäkel, Olaf (2003): Wie Metaphern Wissen schaffen. Die kognitive Metapherntheorie und ihre Anwendung in Modell-Analysen der Diskursbereiche Geistestätigkeit, Wirtschaft, Wissenschaft und Religion. Hamburg: Dr. Kovač.

Kertész, András / Rákosi, Csilla / Csatár, Péter (2012): Data, problems, heuristics, and results in cognitive metaphor research. In: Languages Sciences 34, 712–724.

Majoros, Krisztián (2013): Metapher und Kookkurrenz. Eine alternative “Trichter”-Methode zur korpusbasierten Untersuchung metaphorischer Ausdrücke in öffentlich zugänglichen elektronischen Zeitungskorpora am Beispiel der Wissenschaftsberichterstattung im Bereich der Zellbiologie. In: Sprachtheorie und germanistische Linguistik 23.1, 65–110.

Majoros, Krisztián (2016): Zellenmetaphern und die organische Metaphorik in deutschsprachigen Plenarvorträgen. Die weiterentwickelte „Trichter”-Methode. In: Argumentum 12, 123-151. Available at:

http://argumentum.unideb.hu/2016-anyagok/sondernummer/07_majorosk.pdf  (30.10.2017)

Mio, Jeffery Scott (1997): Metaphor and politics. In: Metaphor and Symbol 12-2, 113–133.

Musolff, Andreas (2004): Metaphor and Political Discourse. Analogical Reasoning in Debates about Europe. Basingstoke: Palgrave-Macmillan.

Musolff, Andreas (2015): Dehumanizing metaphors in UK immigrant debates in press and online media. In: Journal of Language Aggression and Conflict 3-1, 41–56.

Musolff, Andreas (2016): Political Metaphor Analysis. Discourse and Scenarios. London: Bloomsbury. 

O’Brien, Gerald V. (2003): Indigestible food, conquering hordes, and waste materials: Metaphors of immigrants and the early immigration restriction debate in the United States. In: Metaphor and Symbol 18-1, 33–47.

Santa Ana, Otto (1999): ‘Like an animal I was treated’: Anti-immigrant metaphor in US public discourse. In: Discourse and Society 10-2, 191–224.

Schön, Donald A. (1993): Generative metaphor: A perspective on problem-setting in social policy. In: Ortony, Andrew (ed.): Metaphor and Thought. 2^nd and revised version. Cambridge/New York: Cambridge University Press, 137–163.

Schwarz-Friesel, Monika (2013): Sprache und Emotion. 2^nd ed. Tübingen/Basel: Francke.

Schwarz-Friesel, Monika (2015): Metaphern und ihr persuasives Inferenzpotenzial: Konzeptualisierungen des islamistischen Terrorismus nach 9/11 im massenmedialen Diskurs. In: Spieß, Constanze / Köpcke, Klaus-Michael (eds.): Metapher und Metonymie: Theoretische, methodische und empirische Zugänge. Berlin: De Gruyter, 143–160.

Schwarz-Friesel, Monika / Kromminga, Jan-Henning (eds.) (2014): Metaphern der Gewalt. Konzeptualisierungen von Terrorismus in den Medien vor und nach 9/11. Tübingen: Francke.

Schwarz-Friesel, Monika / Reinharz, Jehuda (2013): Die Sprache der Judenfeindschaft im 21. Jahrhundert. Berlin/Boston: de Gruyter.

Semino, Elena (2017): Corpus linguistics and metaphor. In: Dancygier, B. (ed.): The Cambridge Handbook of Cognitive Linguistics. Cambridge: Cambridge University Press, 463–476.

Spieß, Constanze (2017): Vom Flüchtlingsstrom bis hin zum Flüchtlingstsunami? Metaphern als Meinungsbildner. In: Magazin Erwachsenenbildung.at 11. Available at: http://www.erwachsenenbildung.at/magazin/17-31/meb17-31.pdf (30.10.2017)

Stefanowitsch, Anatol (2005): Quantitative Korpuslinguistik und sprachliche Wirklichkeit. In: Solte-Gresser, Christiane et al. (eds.): Von der Wirklichkeit zur Wissenschaft: Aktuelle Forschungsmethoden in den Sprach-, Literatur- und Kulturwissenschaften. Hamburg: LIT-Verlag, 141–155.

Stefanowitsch, Anatol (2006): Corpus-based approaches to metaphor and metonymy. In: Stefanowitsch, Anatol / Gries, Stefan Th. (eds.): Corpus-based Approaches to Metaphor and Metonymy. Berlin/New York: Mouton de Gruyter, 1–16.

Ziem, Alexander (2008): Frames und sprachliches Wissen. Kognitive Aspekte der semantischen Kompetenz. Berlin/New York: de Gruyter.