Metaphors of the virtual: how ordinary people frame what the internet is

This article explores how experiences of the virtual are metaphorically articulated in people ’ s narratives of these experiences. It analyses a database of 171 responses by UK adults of all ages and backgrounds to establish how they used metaphor to frame their experiences online, at a time when the internet was increasingly becoming a platform for living everyday life. The article ﬁ nds that respondents used a multiplicity of metaphoric frames simultaneously, some of them well established and conventionalised in discourse, and some newer. Although there is evidence of a new way of metaphorically framing the virtual as a way of experiencing the everyday, this had not yet replaced conventionalised metaphors in the data, but co-existed with them. The article argues that a slow shift in discourse patterns may be taking place alongside a shift in experience, whereby the role of technology in mediating the everyday is gradually becoming invisible.


Introduction
The language patterns we use in articulating aspects of daily life reveal insights about individual and collective experience. When this experience shifts, conventional discourse around it adapts, and this is particularly the case with the metaphors we use to frame social issues (Lakoff and Johnson 1980). This article explores the ways in which metaphor frames understandings of what the internet is, as these are revealed in the way ordinary people described their online experiences in 2015, in the middle of a decade when virtual technology was not just used for isolated tasks in work and entertainment, but was increasingly becoming all-pervasive in mediating everyday experience.
The article draws on literature about metaphoric framing, to examine a large dataset of written responses to questions around the virtual, given by 171 UK-based respondents of different ages and backgrounds. It uses qualitative metaphor analysis (following Charteris-Black 2011) and quantitative content analysis, and it seeks patterns in language use as a starting point to discuss patterns in social understanding.
The data were collected at a time when smart mobile telephony, portable devices and applications were transforming experiences of the everyday (Deuze 2012), moving the internet from desktop/laptop computers into users' pockets, thus allowing the constant mediation of the everyday through the virtual. The article finds that metaphors which, according to the academic literature reviewed in a subsequent section, had been used in the early days of the popularisation of this technology were far from extinct in the analysed data but, alongside those, a newer metaphoric frame appeared that constructed the internet as an unnoticeable part of life. This has not been previously empirically found within a large dataset of ordinary discourse.
The article argues that scholarly forecasts of a time when the internet will be so integrated into life that metaphoric language will no longer be required to frame it (Markham 2003) could be in the process of materialising, but such a process is slow and evolutionary. Although by 2015 the internet was no longer a new technology, experiencing everyday life through it was still new, and experiences of the virtual remained complex and varied.

Metaphors as frames
Metaphor involves using a word with a sense that is removed from its original meaning (Charteris-Black 2011). It is more than a figure of speech: it plays a role in reflecting and shaping ways of understanding the world. In their cognitive linguistic theory of metaphor, Lakoff and Johnson (1980) proposed that metaphor is not just about using words but about creating links between concepts. It is about constructing a conceptual domain (the target domain) in terms of another (the source domain) and not only transferring vocabulary but also mapping characteristics we associate with the source domain onto the target domain (Lakoff and Johnson 1980). Metaphors assist us in understanding new or complex concepts, by encouraging us to think of them in terms of something we are more familiar with (Mio 1997). Metaphors can also be used ideologically, as part of a rhetorical attempt to persuade, by shaping how an issue is publicly understood (Fairclough 2003;Charteris-Black 2011). Thibodeau, Matlock, and Flusberg (2019) review many of the challenges that have been raised to Lakoff and Johnson's proposal, including questioning whether all or just some properties of a concept get transferred from source to target domain; the extent to which a transfer of vocabulary indicates a transfer of actual mental representations of the concepts concerned; or the need for a better understanding of how these mental representations are formed in people's minds and how they influence any decisions they make. At the same time though, a range of research they review provides evidence that "metaphors can influence how people think about a wide range of sociopolitical issues", even if they don't necessarily predict behaviour (Thibodeau, Matlock, and Flusberg 2019, 9).
An important implication of viewing metaphor as a cognitive structure is that metaphors function as frames. Frames are ways of understanding or discussing an issue which "select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation" (Entman 1993, 52). Frames provide an answer to "what is going on here" (Goffman 1974) by focusing only on some aspects of the event and using them to define the whole event. For example, a political debate may be framed as an exchange of perspectives or as a dispute. Depending on which frame is applied, this carries over different expectations about why the debate is happening, its possible outcomes, or the roles of its participating actors. Therefore, just like metaphors, frames transfer meaning between originally distinct domains of experience.
Not all frames are metaphoric, in that they don't all involve language being used with a non-literal meaning, but some well-studied frames (e.g. the game framesee Aalberg, Strömbäck, and de Vreese 2012) are frames and metaphors at the same time. Burgers, Konijn, and Steen (2016) point out that figurative language, such as metaphor, may fulfil several of the cognitive functions in Entman's (1993) definition of framing mentioned above. Therefore metaphor can be a type of frame in itself, in other words it can function as a metaphoric frame. The term metaphoric frame is used in this article interchangeably with metaphor, because the metaphors explored in its empirical analysis function as frames.
Some metaphors introduced in public discourse to frame an issue may gradually become conventionalised, namely standardised and unnoticed ways of commonly referring to this issue in everyday discourse (Burgers 2016). They may gradually become part of the lexicon, almost literal, settled ways of speaking (Knudsen 2005). The more established and conventional a metaphor becomes as a way of talking and thinking about an issue, the less speakers need to explain it when they use it (Knudsen 2005), the less mental processing they need to understand the connection between its source and target domains (Bowdle and Gentner 2005), and ultimately the more it shapes and reflects how they commonly conceptualise the issue.
However, a conventionalised metaphor does not necessarily remain constant through time. A shift in the conventional metaphors speakers of a language ordinarily use to frame an issue often indicates and results from a modification in their beliefs or experiences of this issue, which are socially and culturally rooted (Jensen 2015;Buaraphan 2011). When a metaphor is no longer the most pertinent tool to express people's current perceptions of an issue, it is either re-interpreted and given modified or new meaning, or it is discarded and replaced by new metaphors with different source domains (Knudsen 2005). Therefore, studying how metaphors change or how new metaphors emerge can provide deeper insights into changing social understandings and experiences of issues (Burgers 2016).

Metaphors of the internet
This section reviews academic literature on the internet metaphors used in popular discourse since the popularisation of this technology in the 1980s and 1990s. The present study is not a longitudinal analysis of these metaphors over past decades, as scholarship already exists about these periods. This article uses the insights delivered by the scholarship reviewed here to analyse its own findings from the 2015 database.
The internet became a heavily framed concept in the early years of its popularisation during the 1980s and 1990s, when it was still a new technology that people were unfamiliar with. Early metaphors helped make sense of the internet in terms of the more familiar everyday domains of place/space, which became essential to understandings of the virtual. These spatial metaphors offered the advantage of conceptualising what is essentially abstract data flows as something relatively stable and localised (Wilken 2007, 51). The information superhighway and the cyberspace metaphors are two instances of spatial metaphors that shaped early thinking about what the internet is by "evoking images of navigation and exploration" (Nunes 1995, 315). The former framed it as a conduit for transient information and the latter as a novel, borderless place (Blavin and Cohen 2002) existing on a "parallel" level to geographical space (Taekke 2002 cited in Wilken 2007, 59). A further early spatial metaphor constructed the internet as a community (Rheingold 1993), focusing on the social rather than the geographical qualities of place (Fernback 2007). By way of continuity, spatial metaphors were also applied to mobile phones when that technology moved the internet into people's pockets (Wilken 2013;Isomursu et al. 2007).
Two decades after the early period of the internet, spatial metaphors persisted in how ordinary people spoke of their experience online (Matlock et al. 2014). An empirical study of 16 university students in the early 2010s found that, although they used a smaller range of spatial vocabulary to describe the internet compared to respondents in the 1990s, these students still conceptualised the virtual as a separate place they "go" to (Matlock et al. 2014).
However, the extent to which the internet, or indeed any mass medium, can be separated from the "real" place of our everyday experience has been questioned. Meyrowitz (1986) suggests that electronic media alter the effect of physical space in our lives by exposing us to real social situations we would otherwise have no access to, and thus they dissociate our social experience and knowledge from our physical location. Deuze (2012) goes even further to argue that, in the early twenty-first century we don't just use more media, nor do we use these media more often, but we live our everyday lives through media and our experience of reality is almost completely mediated: In this space, media have become infinitely intertwined with every single way of being, seeing, moving and actingwithout replacing the world of lived experience.
[…] This is not a life simply lived with more media than before the age of internet and mobile telephony. It is a way of living that fuses life with material and mediated conditions of living in ways that bypass the real or perceived dichotomy between such constituent elements of human existence. (Deuze 2012, 28) wIf, as Deuze states above, the everyday is being increasingly mediated by virtual media, and there is no longer a dichotomy separating real from virtual experience, this poses the question of whether this is reflected in how we talk about this technology. Markham (2003, 3) predicted that over time there would be a "progressive continuum of metaphoric frames" used for the internet: initially conceptualising it as a place and a tool and progressively as a way of being. She pointed out that the internet was originally spoken of and understood as a tool, that enabled people to do things across distances or in speed and quantities that were not possible without it. Then it was seen as a separate place, where people could interact outside their "real" physical environment. However, as the internet would gradually become a seamless part of everyday life, Markham forecast a blurring of the distinction between online and offline and a time when the internet would be "simply the way one learns about, makes sense of, and ultimately knows the social world" (2003,10). This latter metaphor, she suggested, would be linguistically marked by the absence of a particular frame of reference that separates things we do online from things we do in the physical world. In other words, the internet would become a linguistically unmarked way of carrying out aspects of our lives. What Markham foresaw is essentially an evolutionary change in the metaphors used for the internet, whereby "old metaphors are slowly replaced by new ones over relatively long stretches of time" (Burgers 2016, 256). Indeed Markham was not alone in thinking that a day would come when the internet is ordinarily spoken of as real place, as one of the multiple ways of experiencing real life (Nunes 1995;Blavin and Cohen 2002). Carrington (2017) points out that the words online and offline are themselves an outdated metaphor from a time when internet access relied on the fixed location of a telephone landline. She argues that for younger generations there is no longer a distinction between being online and offline and in her qualitative interviews with two young respondents she finds that online and offline states and ways of interaction are intertwined. The present article takes this line of empirical enquiry further, by exploring whether Markham's (2003) prediction is evidenced in the discourse of a large sample of ordinary people in the mid-2010s.
As explained above, both Carrington (2017) and Matlock et al. (2014) explored the extent to which the metaphor of the internet as a separate place remained current by studying relatively small samples of young respondents: 2 and 16 respectively. The former author found a move towards an integration of digital and "real" experience, and the latter a persistence of spatial metaphors even if their linguistic complexity had reduced over time.
The present study addresses this question through a much larger and more diverse sample of UK-based respondents of all ages and different backgrounds, not only young university students as in the above studies. More importantly, this study makes a new contribution to this debate by comparing the relative prominence of not only spatial, but of several different metaphors for the internet in the narratives of these respondents.
This article investigates the extent to which there is any evidence of the "way of being" metaphor in the discourse of ordinary people, and whether there has been a move away from the conventionalised metaphors for the internet reviewed in this section, and towards framing the virtual as simply a way of experiencing the world, as Markham (2003) predicted would happen. This question has not been empirically explored by scholars so far.

Methods
This article analyses a dataset of 171 responses by UK-based adults of all backgrounds and ages to a 2015 survey about their experiences online. As discussed earlier, by the mid-2010s daily uses of the internet in the West had come to fuse the dichotomy between virtual and non-virtual experience (Deuze 2012), so these responses were taken at a time when people's experience of the internet was shifting. Respondents no longer logged on to the internet to complete isolated work and entertainment tasks, but were constantly connected, carrying out most aspects of their personal communication. The dataset is a rich resource to study whether and how this gradual shift in experience may be reflected in metaphor use.
The study first uses qualitative metaphor analysis to identify the metaphors used in the responses to refer to the internet, and then it maps out the relative prominence of these metaphors through quantitative content analysis. It addresses three research questions: RQ1: Which metaphors were used by respondents to frame what the internet is? RQ2: How prominent were different metaphors in the data? RQ3: Is there any evidence of a move away from conventionalised metaphors (place, tool) towards an understanding of the internet as a way of being?
The dataset consists of all the responses to the Mass Observation 2015 "You Online" directive. The Mass Observation Project (MO) is a British national writing panel. Its members are volunteer participants from different geographical and socio-economic backgrounds around the country. Their ages range from 18 to over 90. Three times a year they are sent "directives", namely open-ended qualitative questionnaires on broader themes, covering social and political issues and events, or their personal life experiences. Respondents write their long-form answers in free diary-like style, as "the Project solicits in-depth accounts (both opinion and experience) of everyday life" (Mass Observation Project 2021). Participants are anonymous and identified with a unique code, but their age, gender, occupation and geographical location are available. Their responses to the different directives are stored at the Keep Archive in Sussex, England, where they can be used for academic research.
The data analysed in this article are responses to open ended questions sent to participants as part of the MO's summer 2015 directive, regarding their personal experience of the internet. The questionnaire is available at http://www.massobs.org.uk/massobservation-project-directives and includes questions on respondents' early online experiences, their daily routines, experiences with social networking, the relationship between the internet and their identity, and the relationship between how they experience the internet and other communication media.
The first step in analysing the 171 written responses to the "You Online" directive was to identify which metaphors respondents used to refer to the internet. These emerged inductively from the data through metaphor analysis (Charteris-Black 2011). This qualitative method involves (a) reading through the text of each response to identify all instances where words were deployed with a non-literal meaning within the context of the response, to refer to the target domain of the internet; (b) identifying the original source domains these non-literal words came from; and (c) mapping correspondences between source and target domains (see Charteris-Black 2011). This analysis was carried out following Charteris-Black's (2011) analytical steps above, until a point of saturation was reached and no further metaphors emerged from the data. The researcher also copied typical examples of sentences containing each metaphor from across the data, and these examples are presented in the findings as qualitative evidence to illustrate the different metaphors. The examples are accompanied by the code number of the response they were taken from and they were selected as typical instances of the metaphors discussed. Formal permission to use this material was obtained from the Mass Observation Archive.
A quantitative content analysis of all the responses was subsequently carried out, recording the presence of the different metaphors identified in the metaphor analysis throughout the dataset. The unit of analysis was the overall response. A metaphor was recorded as being present in a response, if the response used words or phrases to describe the internet, whose literal meaning came from that metaphor's source domain. For example, the phrase "to go online" derives from the source domain of physical place (it indicates movement within physical place) and was thus used as an indicator of the place metaphor; whereas the phrase "I use the internet to … " was an indicator of the tool metaphor. Each metaphor was coded on a presence-absence basis (Aalberg, Strömbäck, and de Vreese 2012, 169): I did not measure how many times each response contained a metaphor, but if it contained this metaphor at least once; and each response could contain several different metaphors which were all coded. This was because the purpose of the study was to identify the proportion of people who recognised and used each metaphor (i.e. patterns of metaphor use among the total number of respondents), not how many times each respondent used each metaphor, which may vary by the length of their response or chance.
The responses were also coded for the age and gender of respondents. Table 1 shows the distribution of respondents based on these demographic variables. Fifteen percent of the data (26 responses) were also coded independently by a final year media student. Inter-coder reliability was calculated using Krippendorff's alpha, and it was over 0.78 for each of the metaphors measured.
One limitation of the study is that the questionnaire was not originally designed for the purpose of studying metaphor and thus, as will be discussed in the findings, some of its various open-ended questions contained metaphors of the internet in their own wording. This may have occasionally influenced the respondents in using the same metaphors in their responses. Despite this limitation, as I will discuss in the findings, the data still clearly demonstrate respondents' familiarity with and adoption of these metaphors in their own discourse, as well as their ability to recognise different metaphoric frames and shift between them in their own narratives.
Moreover, the responses also contain other metaphors which were not prompted by the questionnaire, and therefore this dataset provides a very rich resource to explore the emergence of original metaphors in the discourse of people from a range of age groups and backgrounds. As I discussed in an earlier section, previous studies of internet metaphors in ordinary people's discourse have tended to focus on very small samples of teenage respondents or students (Matlock et al. 2014;Carrington 2017) and the present study uses a much broader dataset to provide insights into how people of all ages used metaphor to frame the internet at this crucial time of changing experiences of the virtual.

A multiplicity of metaphoric frames
The first finding of the analysis is that a range of metaphoric frames were used simultaneously by respondents to refer to the internet. There was not a single dominant frame and, as will be explained later, conventionalised and newer metaphors coexisted within the same responses. This suggests that, like many other concepts (Kovecses 2016), the internet was metaphorically understood in multiple ways at the same time and, although non-conventionalised metaphors had entered the everyday lexicon (as this lexicon was manifested in the responses), these had not yet replaced established metaphors like those of the internet as a place or a tool. At the time of the survey, everyday experience was increasingly being mediated by the virtual (Deuze 2012) and the metaphor which frames the internet as a way of being (Markham 2003) was indeed evidenced in the data. However, it had not replaced ways of understanding the internet as something which is separate from daily life. In other words, there is evidence of a non-conventionalised metaphor having entered ordinary discourse, but not of fundamental metaphor change (Burgers 2016).
Conventionalised tool and place metaphors were present in almost all of the responses. 95% of responses contained at least one instance of the tool metaphor and 93% had a spatial metaphor, making these the two most used metaphors in the analysed material (Table 2). This might have been influenced to an extent by the phrasing of the MO directive's own questions: the verb "use" appears ten times in combination with the internet or websites in the questions, thus potentially encouraging respondents to deploy the tool metaphor (examples of these questions include "do you use the internet?" and "what do you use [social networking sites] for?"). Similarly spatial metaphors appear six times in different questions (e.g. "what sites were important to you when you explored the online world?").
Moreover, the directive's questions include two more metaphors, the internet as a human agent (e.g. "the internet is only 25 years old, how do you think it will influence society and personal identity over the next quarter century?") and as a substance ("do you ration your Internet usage?"). These metaphors also appear in 71% and 31% of the responses respectively (Table 2). This makes it difficult to judge if the respondents would have used these metaphors as regularly as they did, had they not been prompted. However, all these metaphors also emerged in the data without any prompting (e.g. "the Net told us the firework cruise was full": a human agent metaphor used as part of an account of the respondent's early experiences with the internet).
Irrespective of the wording of the questions though, respondents' use of these metaphors still tells us two important things, which were not influenced by the questionnaire: (a) that respondents comfortably move between different metaphors when they refer to the internet; and (b) that they still recognise, understand and use tool and place metaphors in their discourse without questioning them, so these metaphors are not extinct.
Respondents appear to see no conflict between using place, tool, agent and other metaphors within the same paragraphs or even the same sentences in their responses. Indeed only 6 responses (3.5%) contained just one metaphor, and these responses tended to be shorter in length. All the other respondents used two or more different metaphors, while 56.7% of respondents used 4 or more metaphors (Table 3). This suggests that different ways of framing the internet are complementary, as each metaphor puts emphasis on some aspects of the technology and obscures other aspects. Moreover, even if it had been the case that respondents used some metaphors more often than they would have if they had not been prompted by the questions, the fact that they recognised and used them suggests that tool and place metaphors were still current and made sense to them. If place and tool metaphors had been abandoned in everyday speech, respondents would have questioned these metaphors and not replicated them in responding.

The internet as a way of being
A second key finding of the analysis is that the metaphor of the internet as a way of being emerged spontaneously in 64% of the responses (Table 2). This happened without any prompts in the questions. This suggests a shift had taken place in discourse which rendered the role of the internet in mediating everyday tasks unnoticeable and thus unmarked in these responses.
Alongside other ways of understanding the internet, respondents to this survey had also come to see it as embedded in how they go about everyday activities like entertainment, shopping and interacting with others. In the following examples, its presence as a medium was not articulated in how respondents described these activities: "Those witty things you shout at the TV you can now shout to the world" [B5567] "I had an ectopic pregnancy two years ago and being able to share my feelings and experiences with other women was a lifeline" [C3210] "I'm not that interesting, so if some shady government official wants to waste their time snooping on me, they're more than welcome" [E5551] "I worry because of my dementia that I will be taken advantage of" The above are some typical excerpts where respondents deployed the metaphor of the internet as a way of being (the words expressing the metaphor are in bold). Respondents referred to activities they or others engaged in online as if these happened in the physical world, without mentioning the internet as the locus or mediator of these actions. The activities they mentioned ranged from interacting with others ("shout", "share my feelings", "debating a point"), to solicited ("I can track my husband") or unsolicited surveillance ("officials [..] snooping on me"), potentially dangerous behaviour ("be taken advantage of") and everyday errands ("shopping", "reported a street light failure"). These are instances of metaphoric language. The literal meanings of the verbs highlighted in these examples do not involve a mediating platform like the internet. Their definitions by the Oxford English Dictionary (OED) are cited below, alongside the first recorded use of these original meanings. All definitions describe unmediated physical action or speech: Shout (c1385): "To utter a loud call, to make a loud outcry expressive of joy, exultation, etc. or to raise an alarm" Share (1566): "To cut (something) into pieces; to break (something) apart by cutting; to cut up, divide." Snoop (1832): "To move in a sly or prying manner." Take advantage of (1393): "To avail oneself of a person or thing. Frequently in negative sense: to seize an opportunity of unfairly profiting by a person or thing, esp. sexually." Debate (1530): "To engage in discussion or argument; esp. in a public assembly" Shop (1806): "To buy or view goods in a shop" Buy (c1000): "To get possession of by giving an equivalent, usually in money" Track (1565): "To follow up the track or footsteps of" Report (1402): "To give an account of (a fact, event, etc.); to relate, recount, tell".
The source domain of all these verbs is that of physical action or speech as manifested in everyday life, or "a way of being" in Markham's (2003) words. The internet is not marked as the mediator of any of these experiences. The use of the metaphor of the internet as a way of being was not restricted to any particular age group. No corelation was found between the respondent's age and their use of this, or indeed any of the other metaphors. Common sense might suggest that younger people, who grew up with the virtual integrated into their lives, would be more likely to see no distinction between online and offline (Carrington 2017) than older people who might not be equally engaged with digital technologies. However the data analysed here suggest that this is not necessarily the case, at least to the extent that the metaphor of the internet as a way of being may indicate how these respondents experienced the virtual.
Indeed, there was only one occasion in the data where a respondent questioned the wording of the discourse in the directive's questions, and this was one of the older respondents, at 75 years old: I've just read this directive again and realise that whoever wrote it is a reluctant old Luddite dragged into the twenty-first century, yet the internet is now part of my life and I appreciate it. [P3209] The above respondent reiterates that he sees the internet as part of his life, or as a way of being. Importantly though, he seems to identify a clash between his own experience and that represented by the discourse adopted in the directive's questions, thus implying that the questions do not portray the internet as part of everyday life. In this respondent's view, the discourse of the questions shows "reluctance" to embrace the internet.
The Luddite metaphor he uses is a deliberate invitation for the reader of his response "to step outside the dominant target domain[s] of the discourse [which was used in the directive] and look at it from an alien source domain" (Steen 2011, 37). This metaphor makes a connection between the questionnaire's authors and the source domain of the nineteenth century English textile workers association, who resisted the introduction of textile machinery in their trade; the implication being that the authors of the directive are failing to discursively acknowledge the centrality of the internet in twenty-first century life, partly through their repeated uses of conventional tool, place and agent source domains. Just like machinery became an embedded and unnoticed part of doing textile work in the nineteenth century, the internet is here rhetorically constructed as embedded in twenty-first century living.
The internet as a metaphysical entity, war zone and resource Three further metaphors were identified in the respondents' discourse, but received considerably fewer mentions than the ones discussed so far: the internet as a (threatening) metaphysical entity or monster; as a war zone, and as a valuable marketable resource. Their use in combination with the metaphors discussed so far indicates that for these respondents the internet has a range of qualities which need to be expressed through diverse metaphors. The metaphors in this section, however, put less emphasis on the function the internet may have in everyday life (e.g. a place, tool, or agent in people's lives) and they draw attention instead to the potential risks and the commercial aspects of the technology.
The metaphor of the internet as a resource appeared in 5% of responses (Table 2). It frames the internet as a commodity that has a marketable value and may or may not be in good supply: "I simply couldn't afford the internet" [H4611] "It's so cheap nowadays, it's not a worry to use it quite a lot" [P3059] The emphasis of this metaphor is on the internet as a paid-for commercial service rather than as a communication medium. Both "afford" ("to be in a position to offer or purchase something" according to its OED definition) and "cheap" ("that may be bought at small cost") derive from the domain of physical resources and here the internet itself is constructed as something that can be bought. The internet as a metaphysical entity (3% of responses, Table 2), on the other hand, uses imagery from mythology to construct the virtual as other-worldly and potentially dangerous: "Sadly now the genie of social media is out of the bottle, it can never be put back" [H2637] "I am deeply suspicious of the internet although now I feel that as [with] Pandora's box, we all have to be as careful as possible" [I1610] "And now science fiction is the norm, what can become more fantastical than that? […] Mary Shelley 'made' a human." [A1706] In the above examples associations from mythology (Aladin's genie, Pandora's box) and literature (Frankenstein's monster) are evoked to frame the internet as uncontrollable and potentially riskythese features are shared by all three mythical entities in the source domain. The metaphor also frames the internet as something outside ordinary experience and thus mysterious and unknown.
Finally, the internet as a war zone metaphor (3% of responses, Table 2) represents it as a setting for battle, usually against criminals: "This is particularly relevant in the fight against terrorism" [N5212] War metaphors are common in policymakers' discourse around online regulation and are used to support measures to control online activity (Dekavalla 2021). Like the other metaphors in this section however, the war metaphor does not appear very often in the data, suggesting that this is not primarily how the internet is viewed by users. Although respondents were explicitly asked about online risk in the questions, very few of them constructed online risk as an enemy to be fought against.

Conclusion
This article empirically assessed scholarly forecasts of a shift in metaphors used to refer to the internet, away from spatial and tool metaphors and towards a conceptualisation of the internet as a "way of being" (Markham 2003). It did this by analysing discourse about the internet from 2015, the middle of a decade when experience of the everyday was increasingly mediated by smart portable technology and constant connection to the virtual (Deuze 2012). The article found that such a metaphor change had not yet taken place in the UK-based survey responses it analysed, as metaphors which scholarly research had identified in previous decades remained prominent. However, there was also evidence of the "way of being" metaphor emerging spontaneously in these responses, and being used alongside old conventional metaphors.
This may indicate a period of transition, when many different metaphors co-exist to express people's diverse experiences of the virtual. The internet may have no longer been new in the mid-2010s, but the experience of living most aspects of daily life through it was still novel. Conventionalised metaphors can take long to adapt, be negotiated and fundamentally change or be replaced in response to slow changes in people's experience or their circumstances (Jensen 2015). Respondents to the 2015 survey recognised a range of different aspects of the internet, which were all still relevant and were foregrounded by the different metaphoric frames they used. Tool and spatial metaphors, which first emerged in the early years of virtual technology (Blavin and Cohen 2002), were deployed alongside a framing of the internet as an unnoticed way in which people experience the world.
These findings don't necessarily contradict the argument that in the digital era we all live our lives in media (Deuze 2012). Instead they qualify this observation, by suggesting that this has perhaps not been happening for long enough to replace well-established metaphors of the internet. Ultimately metaphor helps us understand the new and unfamiliar in terms of the old and familiar (Mio 1997), a kind of read-view mirror to interpret new ways of experience in light of the comfort of what we know well (McLuhan 1964).
How we use metaphor in daily discourse matters because "different linguistically based frames of reference can have meaningful consequences on action" (Markham 2003, 15). How we speak about a concept is connected to how we think about it. Although the connection between this and how we eventually act is complex and also impacted by social factors, patterns of thought and speech may influence what lines of action we perceive as being available for us to choose between. Frames are powerful because they help us make sense of the world around us, at the same time as restricting where we focus our attention when we look at this world (Goffman 1974). If we cease to notice the presence of technology in performing our everyday tasks, we may no longer notice the power it has to shape the experiences that it mediates.
The findings of this analysis are particularly relevant in a post-pandemic world. The global Covid pandemic which started in 2020 and the confinement measures taken by national governments to contain it accelerated the experiencing of daily life almost exclusively through virtual technology for a period. Although, at the time of writing, pandemic lockdowns had been lifted in many countries, the impact of this period on how we work, confer and come together socially is still unfolding. For instance, we speak of attending work meetings without actually presenting ourselves physically in a work setting, and without necessarily qualifying these meetings as "online". Post-pandemic discourse is thus likely to have shifted even more towards the "way of being" metaphor discussed in this article, as our experience of the everyday shifted even further, although this would require further empirical evidence to establish.