An Ethnography of an Imaginary Road: Fear, Death, and Storytelling in the Icelandic Westfjords

Abstract This article presents a historical ethnography of an imaginary road. Drawing on printed sources, archival material, and new field research, the article analyses the Icelandic folktale of ‘Loss of Men on Heiðarbæjarheiði’ (Manntjónið á Heiðarbæjarheiði), a story of regional importance in the Strandir district of the Icelandic Westfjords, especially the fjord of Steingrímsfjörður. The article shows the shape this story takes when encountered locally, where it appears in the form of minimalist place-storytelling that is actualized in the engagement with particular places. It thus contributes to our understanding of how legends work ‘on the ground’. In this local form as place-storytelling, the narrative shows considerable variation and a strong focus on the interpretation of local place names. Based on the contexts and variation observed between the different variants of the story, this article reads ‘Loss of Men’ as a formulation of collective fears, thereby contributing not only to research on legends, but also to the current discourse on emotions in the humanities more generally.


Introduction
This article presents a historical ethnography of an imaginary road: the road over the mountain Heiðarbaejarheiði, which is imagined, once upon a time, to have connected the Strandir region of the Icelandic Westfjords to the rich fishing grounds of Breiðafj€ orður. Historically, this road never existed; it makes its only appearance in the Strandir folktale of Manntj onið a Heiðarbaejarheiði ('Loss of Men on Heiðarbaejarheiði') and place-lore connected with this story. Nevertheless, rich material from throughout the twentieth and into the twenty-first century suggests that this story was and is very widely known in the Strandir district; unlike many highly localized tales, 'Loss of Men' is a story of regional importance. The many different incarnations of this story, a striking number of which were formulated by local inhabitants in local venues without the involvement of academic collectors, allow us a fascinating glimpse of the normal modes of existence of this tale. This glimpse is highly informative not only because it stands in marked contrast to the presentation of the story in its bestknown-and indeed only widely accessible-publication in Guðni J onsson's twelvevolume collection Islenskir sagnapaettir og pj oðs€ ogur (Icelandic legends and folktales) 8: 71-75), 1 but, more importantly, because a study of this legend can grant us insights into how legends work. The rich and richly varied attestations of 'Loss of Men' make clear to what degree the still-living oral tradition in the area works separately from the written accounts even while being tightly intertwined with them. It illustrates how common motifs can move between narratives, which can mean that even the primary actors of a story can be exchanged for different persons without affecting the main thrust of the tale. Also, the material connected with this legend even shows how, given the presence of the right environmental catalysts, one narrative can lead to the creation of another, finally culminating in the emergence of a whole imaginary road. 'Loss of Men' thus can serve as a laboratory case for how oral narratives work in connection with their environment and, in this way, make a contribution to the wider discourse on the legend genre (for instance, Gunnell 2014Gunnell , 2018D egh 2001;Tangherlini 1990; on legends related to landscape, Gunnell 2009Gunnell , 2008. Another discourse for which 'Loss of Men' can be highly relevant is the study of emotions. In its home environment in Strandir, one meets the story in the form of a plethora of short belief statements connected to specific places and their names. These belief statements show considerable variation, and considering what is constant and what varies in them can throw a tantalizing spotlight on what seems to be a mechanism closely comparable to Aby Warburg's theory of images, which he interpreted as results of a 'phobic reflex' (B€ ohme 1997, 144-46)-a way of dealing with fear by giving it shape. 'Loss of Men' is the story of an imaginary disaster that occurred on an imaginary road, which seems to put collective fears both into words and into the landscape. This article thus presents a detailed case study of how collective fears are tackled by landscape-related storytelling. In doing so, it contributes to the wider discourse on emotions that has come to the fore with the recent 'emotive turn'. The study of fear plays an important role in this field, as exemplified by Debbie Felton's recent Landscapes of Dread in Classical Antiquity (2018) or Yi-Fu Tuan's classic Landscapes of Fear (2013). In this context, the material analysed here stands out because it addresses fears caused by threats in the more general landscape rather than particular embodied agents (such as a specific dangerous or monstrous entity), which appear to dominate other discussions.
This article studies 'Loss of Men' by drawing on a broad spectrum of sources that include Guðni J onsson's classic published version, mentions in local publications from Strandir going back almost a century, and various descriptions of local farms. The latter were composed partly by local inhabitants on their own initiative, and partly resulted from interviews conducted by the Place Name Institute of the Icelandic National Museum ( € Ornefnastofnun Þj oðminjasafns) in the second half of the twentieth century. All of this is supplemented by local archival material and personal fieldwork. I will approach these data not only through analyses of the textual sources as texts, but also by relating the various materials to the specifics of the local topography, climate, work practices, broader patterns in the Icelandic storytelling tradition, and geographical distribution of the attestations. Bringing together all of these perspectives will then lead to an overall interpretation of the imaginary road over Heiðarbaejarheiði as a crystallization of collective fears.

'Loss of Men on Heiðarbaejarheiði'
The fjord of Steingr ımsfj€ orður is framed by mountain ranges, the highest of which is Heiðarbaejarheiði. Heiðarbaejarheiði forms a long, narrow ridge that for some two dozen kilometres runs in a straight line oriented roughly north-west to south-east; its eastern side is dominated by steep cliffs (Figure 1; Map 1). Heiðarbaejarheiði rises to a height of some six hundred metres; this makes it tower head and shoulders over the surrounding ridges, whose height does not exceed four to five hundred metres.
Iceland's low population density and extreme topography and climate have always made road-building an enormous challenge. Well into the twentieth century, the easiest way to get from Steingr ımsfj€ orður to the southern parts of Iceland, especially to the rich fishing grounds off Snaefellsj€ okull in Snaefellsnes, was by crossing the mountains on footpaths and bridle-paths that, at best, were marked by cairns erected by stacking up local stone to form man-high drystone cones. During winter in particular, travelling on those paths could be highly hazardous. Travellers avoided the valley floors, as the snow accumulated there made the going hard, while avalanches posed a constant threat. Where at all possible, the preferred option was to stick to the tops of ridges, which were kept reasonably snow-free by the wind and were above the avalanche areas. At the same time, however, these ridges are the most exposed parts of the landscape; if a sudden change in the weather occurred, their shelterless expanses could become deathtraps in their own right, and a multitude of short anecdotes, as well as longer stories, mention people dying from exposure because they were caught out by the weather. However, a tradition relates that this upland was much travelled-through in winter in past times by the people from Steingr ımsfj€ orður and other people from Strandir, when they went on fishing trips to the fishing stations around Snaefellsj€ okull. Now shall be told of one of these mountain crossings. It should be emphasized that learned and knowledgeable people there in the nearby communities say that they do not know in which century these events happened that the story tells of, because they are mentioned neither in annals nor in the Yearbooks of J on Esp ol ın. But the many place names that are known and that according to the story are connected with accidents and misadventures that must have happened there, point decidedly towards the story being true in its main features. And it goes as follows: In the seventeenth century or early in the eighteenth it happened that a short time after New Year eighteen men from the northern part of the Strandir district wanted to go on a fishing expedition westwards under Snaefellsj€ okull, as happened then very often. They intended to go over Heiðarbaejarheiði and set out early in the morning, probably before or around dawn, most likely from the farm Heiðarbaer, which is the farm closest to the upland on its northern side. It was said that these were men of vigour in their prime, who were no hotheads, as one says. When the day was a bit advanced, murderous weather broke loose from the north, a snowstorm with fierce frost and heavy snowfall. Probably these men were carrying a great amount of baggage with them, such as provisions for fishing and clothing, even though the story does not explicitly mention that. That must probably have made the journey hard for them.
Nothing else is then told about the journey or the death of these companions than this: that the bodies of some of them were found at the places which since then are named after them. In truth there are traditions about that-that in the evening of the same day two women were outside at the front door of the farm Gr oustaðir with a light. Then a man, covered in snow, arrived and squeezed in through the farm gate. The women asked him for his name, and he said: 'I was a man,'-and he then rushed out again into the snowstorm, and nothing was heard of him afterwards. It is uncertain whether grown-up men were then present at Gr oustaðir, 4 and it is not mentioned that he was pursued or that anyone searched for him. It was said that there was ice on Gilsfj€ orður fjord, and then it was said as most likely, that he had run onto the ice during the snowstorm and there perished in some manner. It was also said that this man was one of the eighteen companions who on that morning had set out for Heiðarbaejarheiði.
It seems likely that most of these companions turned back on the upland and wanted to attempt to reach the settled area north of the mountain. How long this dangerous weather lasted does not emerge from the story, nor whether a search was undertaken after the storm, but it is said that the bodies of eleven of these companions were found at the places which since have been named after them. B€ oðvar was found by the so-called B€ oðvarslaekir ('B€ oðvar's Brooks'), furthest to the north on Tunguheiði [Tr€ ollatunguheiði]. Ing olfur was found in Ing olfsl ag ('Ing olfur's Hollow'), a little uphill from Tr€ ollatunga. H akon was found by H akonarlaekur ('H akon's Brook'); that flows over the bottom part of the home-field of Tr€ ollatunga and eastwards into the river Tungu a. Thus it seems as if these men had been on the right way home to Tr€ ollatunga, when they gave up, and they are remembered only by the hard section of the trail. Hr olfur was found in Hr olfsm yri ('Hr olfur's Wetland'), which is between Tungugr€ of and H usav ık, below Tr€ ollatunga. Up in the valley Tungudalur the bodies of five men were found. Bjarni was found in Bjarnagil ('Bjarni's Glen'; Figure 2). He had fallen to his death there. B arður was found on the so-called B arðarbreiður ('B arður's Broads'), J on on J onsv€ orðuhjalli ('Rock Terrace of J on's Cairn'), and two brothers on Braeðrav€ orðuhjalli ('Rock Terrace of the Cairn of the Brothers'). It is said that one of them was called Narfi, and he was found on so-called Narfaengi ('Narfi's Meadow'), which is down in the valley Miðdalur on the land of the farm Gestsstaðir. Some say that Isleifur, one of these companions, was found on Isleifsm oar (' Isleifur's Moorlands'), very close to Heiðarbaer,-that would then be the eleventh one-, but some say that he wasn't one of the companions, rather he died there later. Then it would be almost certain for eleven or twelve of these companions, what fate befell them. About the six or seven, who were not found, it has mostly been suspected that they will have fallen from the edges of cliffs, especially in Miðdalur valley or in Hraundalur valley, and will have suffered death in this manner, because later on no trace was ever found of any of them. About their names also nothing is mentioned.
Nothing about this story seems unlikely, even if all more precise pieces of evidence are lacking, such as the time period when the story happened, the place of origin of the men who died, etc. It seems likely that all or most of these men were from the communities around Steingr ımsfj€ orður fjord, the communities Kirkjub olshreppur, Hr ofbergshreppur, and Kaldrananeshreppur, and if it was like that, also some from Arneshreppur. It is obvious that a great loss happened at many places, and probably many a wound [from grief] that never healed until death.
In all those details that are of significance for the matter, this story is recorded after a manuscript of G ısli J onatansson in Naustav ık near Heydals a in the district of Strandir. Even in its published form, this narrative shows the marks of a circulation which is both complex and deeply local. Guðni J onsson bases his text on a manuscript by P etur J onsson of Stakkur, who was a teacher in Strandir. P etur himself based his version on a manuscript he had received from G ısli J onatansson (1904-92), who lived on the farm of Naustav ık, and G ısli J onatansson refers to the opinions of 'learned and knowledgeable people there in the nearby communities' ('fr oðir og kunnugir menn par ı n alaegum byggðarl€ ogum'): in this story, we are not dealing with a single author's narrative, but with a written-down composite version of oral traditions that were in wide circulation. And while the way the story of 'Loss of Men' circulated in this tradition appears to have been rather complex, at least judging from Guðni J onsson's account, this circulation also seems to have had a strong regional focus: P etur J onsson probably had got hold of the story while working as a teacher in the Strandir area; G ısli J onatansson's farm Naustav ık is, as the crow flies, only about four kilometres distant from Heiðarbaejarheiði; and the oral sources that he refers to are explicitly described as local people. In its transmission, 'Loss of Men' very much appears to be a story of the south coast of Steingr ımsfj€ orður. These deep local roots of the story tally with the emphasis that the narrative itself puts on place names. Early on, the narrator makes a key statement (Guðni J onsson 1940-57, 8: 72): En hin m€ orgu € ornefni, sem pekkt eru og sagan telur standa ı sambandi við atburði p a og slysfarir, sem p a hafi att að hafa gerzt, benda eindregið til pess, að sagan muni vera s€ onn ı aðaldr attum.
But the many place names that are known and that according to the story are connected with accidents and misadventures that must have happened there, point decidedly towards the story being true in its main features.
In this remark, place names are invoked as the main testimony to the truth of the story. This is particularly important given the openness with which the narrator admits to how problematic the transmission of the story is otherwise: the narrator himself emphasizes in the sentences immediately preceding this reference to place names that the incident is not mentioned in any written historical sources and that the time when it is supposed to have happened is unknown even to well-informed local people. (That the following paragraph dates the events to the seventeenth or eighteenth century appears to be, in light of this admission of ignorance, an educated guess.) In the following, this emphasis on place names forms a leitmotif running through the whole narrative. The narrator states repeatedly that the bodies of the victims were discovered at those places which henceforth were named after them (Guðni J onsson 1940-57, 8: 73 [twice]), and even states clearly that nothing else is known about their misadventures beyond the connection between place names and the recovery of their bodies: Ekkert segir s ıðan af ferðalagi eður afdrifum peirra f elaga annað en pað, að l ık nokkurra peirra fundust a peim st€ oðum, sem s ıðan eru við p a kennd.
Nothing else is then told about the journey or the death of these companions than this, that the bodies of some of them were found at the places which since then are named after them.
(emphasis added) To flesh out the gaps in the skeleton structure created by the names of dead men, the narrator every now and again inserts his own speculations. Thus, early on in the story, he suggests: 'Probably ("sennilega") these men were carrying a great amount of baggage with them, such as provisions for fishing and clothing, even though the story does not explicitly mention that ("p ott sagan geti pess ekki s erstaklega")' (emphases added). Later, he states, referring to general probability: 'It seems likely ("l ıklegt pykir") that most of these companions turned back on the upland.' It is worth highlighting how open the narrator is about the gaps in his knowledge. He presents a coherent, step-by-step narrative, but at the same time he makes a clear distinction between what he knows and what he guesses. What the narrator knows are the place names, and what he guesses is more or less everything else; the little incident at Gr oustaðir is pretty much the only element of the story that is based on a tradition which is not toponymic, and thus shows social circulation of an overarching narrative. Apart from this single small incident, the place names provide the anchor and framework for a narrative that in almost everything but its references to place names admits its own insecurity. 'Loss of Men' illustrates that this way of thinking about place names was not a quirk of an individual author, but constitutes a broader cultural habitus, and indeed one which in Iceland has been well-established since the Middle Ages (see, for instance, Þ orhallur Vilmundarson 1991). It has already been mentioned that the printed version of the story published by Guðni J onsson explicitly addresses its complex transmission and its embeddedness in the collective discussions of 'learned and knowledgeable people there in the nearby communities'. If the presence of the story in Strandir is considered with reference to a broader selection of material, the deep-rootedness of 'Loss of Men' in the narrative traditions of Steingr ımsfj€ orður becomes even clearer.

'Loss of Men on Heiðarbaejarheiði' as Found in the Strandir Area
In Strandir a long tradition exists of engaging with place names and the storytelling related to them. Large-scale recording of this storytelling tradition started in the nineteenth century at the latest and originated from a broad range of social contexts: some work was done by schoolteachers, some by academic institutions of the Icelandic state, and other stories were written down and circulated directly by the farmers whose land they touched upon. Guðni J onsson's version of 'Loss of Men' is a good example of the dynamics involved, as Guðni had not himself 'collected' this story, but rather received it from a local teacher, P etur J onsson; and P etur in turn had done little but pass on what he had received already in a tidy manuscript from a local farmer, G ısli J onatansson of Naustav ık. The local farmer was the primary actor, and academics entered the stage only much later.
In fact, the first written recording of 'Loss of Men' had nothing much to do with academia. In 1929, Filippus M. Gunnlaugsson (1905-81) wrote down a version of the story for the periodical Viljinn (The will), 5 a monthly journal edited by the Young Men's Association Geislinn. It was produced only during the winter months, when agricultural work did not take up as much time; since printing was not available, the copies of Viljinn were written by hand and then passed from farm to farm. To this journal, Filippus contributed a description of the farm of Os. In this text, he gives a systematic account of the place names found on the land of the farm-or at least of the 134 toponyms that he thought most important-which also included a selection of the stories and explanations connected with these names. In a manner very typical of descriptions of farms published during this period, these explanations range from the workaday to the fantastical. Thus, St orahl ıð ('Big Slope') is simply 'a very big area of pasture' ('engjaflaemi allst ort'; p. 1); T ıkarsund ('Bitch's Defile') is so called because a bitch that accompanied the workers on the pasture once had her puppies there (3); Gvendarengi ('Guðmundur's Meadow') had this name ever since an accident in which a boy of that name drowned in a waterhole (4); and Dvergahlaði ('Pile of the Dwarfs') is the place where dwarfs once started building a bridge across the fjord, though they soon abandoned the enterprise, leaving behind only a half-finished bridgehead, identified with a lava formation which can be seen to this day (7). Filippus's description also covers the two places Þorsteinssund ('Þorsteinn's Defile') and That was on one occasion, when Heiðarbaejarheiði was a much-used route, that eighteen mensome say that they were schoolboys-during wintertime undertook to travel on it, and they were caught up in a whiteout. They got lost and separated from each other and all died of exposure. One of them, however, managed to get into the farm gate at Gr oustaðir in Gilsfj€ orður, but he had lost his mind by then and ran back out into the snowstorm, and was then found dead not far from there. Most of the others were found in the summer, and among them one who was called Þorsteinn, and he was found right at the front part of Þorsteinssund, which since is named from him. This account is interesting in several respects. For one, it is a good example of the format in which one normally meets 'Loss of Men'. Typically, where I have encountered the story, this encounter takes the shape of very short, matter-of-fact statements that place such-and-such is named from the man so-and-so who was one of those who died when they tried to cross Heiðarbaejarheiði. A broad range of examples is found in the files of the Place Name Institute of the Icelandic National Museum ( € Ornefnastofnun Þj oðminjasafns), which operated from 1969 to 1998. 6 Most of these files for the Strandir area were collected in the 1970s by interviewing local informants, typically people who had worked on local farms for several decades, were already of very advanced age, and were asked to give descriptions of their farms, the place names on these farms, and the stories connected with them. In this corpus of material, another example of the typical tendency to brevity is found in an account of the farm of Tindur, which today is abandoned: 7 Runatagl. Helstu l ıkur fyrir pv ı nafni er, að parna hafi maður orðið uti, en nokkrir menn voru a ferð h er suður yfir fjallgarðinn og t yndu allir l ıfi, fundust v ıða, baeði h er norðan fjalls og fyrir sunnan heiði. Maðurinn hefur p a heitið Run olfur.
Runatagl ('Runi's Tail'). The greatest likelihood for that name is that a man died from exposure there, when several men were on a journey here towards the south over the mountain massif and all lost their lives, were found all around, both here north of the mountain and south of the upland. The man was then called Run olfur.
Equally laconic is a description of the farm of Tungugr€ of, which until its abandonment was the farm neighbouring H usav ık: 9 . . . . Tyrfingshvammur ('Tyrfingur's Grassy Hollow'). No turf or peat was cut there. Guðj on supposes that the name is derived from a male personal name. On these paths there are in many places toponyms derived from male personal names, and it is said that they are named from a group of men, probably fishermen, who lost their way on Tr€ ollatunguheiði (or Heiðarbaejarheiði, see the place name account of Tr€ ollatunga) and died of exposure.
This Tyrfingshvammur is another addition to the list of places connected with the story. Here, furthermore, a certain fluidity of the tradition not only becomes tangible in the reference to a place name that is not found in other accounts of the story, but also in how the place name is handled and in the cavalier approach that the description takes to otherwise well-established details. In this account, the handling of the place name itself is explicitly highlighted as conjectural: Guðj on explicitly 'supposes' (gizkar) that the toponym belongs to the context of 'Loss of Men'. This appears to be an inference based on etymological analysis rather than on the narrative itself. The account of Tyrfingshvammur starts with the statement: 'No turf or peat was cut there'. This appears to reflect a musing on etymology. Tyrfings-is reminiscent of the Icelandic word for turf (torf). So here-or so it seems-we get a glimpse of the informant thinking about the background of the toponym through its meaning. The name sounds a bit like turf-cutting, but then Guðj on knew that turf was never cut at that place, so the name could not have been derived from turfcutting, and therefore it seemed more likely to him that it was just another of the many place names derived from 'Loss of Men'. Yet Guðj on was not sure; what he voiced is just a supposition (gizkar), and his uncertainty may also be reflected in how he deals with an otherwise consistent element of the story: in this account, the 'Loss of Men' occurred on Tr€ ollatunguheiði, the next range to the west of Heiðarbaejarheiði. Heiðarbaejarheiði itself is only mentioned in an insert in brackets which may be due to Guðj on himself, but sounds more like an editorial insert made by the interviewer, who noted and highlighted a discrepancy with how the story appears in other files.
The hollow to the south of Torfskarð is called Ing olfsl ag ('Ing olfur's Hollow'). The story tells that it, together with several place names derived from personal names on these paths, is named from schoolboys from H olar (some say fishermen), who died of exposure at these places, lost on the way along Heiðarbaejarheiði . . .
In this account, the fluidity of tradition surfaces when the victims of the fateful trek over the mountain become pupils of the cathedral school at H olar. In almost all of the passages quoted, the dead men had been fishermen. Filippus M. Gunnlaugsson had mentioned an alternative tradition ('some say') that saw them as schoolboys; here they become associated with one of Iceland's most famous schools (see Einar Olafur Sveinsson 2003, 206), and it is their identification as fishermen which is downgraded to 'some say'. 'Loss of Men' can be treated as a rather loose framework into which place names can be inserted quite freely and where details can shift to the point where almost the only stable element is the basic argument: 'place name X is derived from a male personal name; therefore, here the body of one of the men who died of exposure in the mountains must have been found'. If one tries to map the toponyms that got entangled in the 'Loss of Men' tradition, it soon becomes apparent over what a remarkable geographical area they are spread out (Map 1): some of the names derived from 'Loss of Men' are located at a surprising distance from Heiðarbaejarheiði. This too is illustrated best by Filippus: Þorsteinssund and Þorsteinsfell, where he places the body of one of the victims of 'Loss of Men', are located more than a dozen kilometres north of Heiðarbaejarheiði as the crow flies. The idea that anybody would be able to cover this distance in whiteout conditions is hardly believable-especially considering that the route from Heiðarbaejarheiði to Þorsteinssund and Þorsteinsfell would run crosswise to the predominant direction of the mountain ridges on the south coast of Steingr ımsfj€ orður; that is, in order to get from Heiðarbaejarheiði to Þorsteinssund, one would have to climb up and down some five mountain ridges (Map 1).
Filippus's narrative is not the only one which connects toponyms with 'Loss of Men' that are strikingly far away from Heiðarbaejarheiði. Not much closer than the place names Filippus talks about are some on the land of the farm Þiðriksvellir, connected with a variant of 'Loss of Men' in a description of the farm from the 1950s (Figure 3): 15 Þ a s€ ogus€ ogn hefi eg heyrt, að vermenn, er komu sunnan ur Barðastrandars yslu og aetluðu til r oðra að Gj€ ogri ı Strandas yslu, hafi hreppt vont veður a heiðinni, villzt og ekki komizt til byggða. Hafi peir allir l atið l ıfið og sumir findizt ı Halld orshvammi og P alshvammi. Eru parna v ıða dj up og hrikaleg glj ufur, sem haettuleg geta talizt. Halld orshvammur ('Halld or's Grassy Hollow') and P alshvammur ('P all's Grassy Hollow') are located some ten kilometres, as the crow flies, from Heiðarbaejarheiðiin order to try to cross the one and to die at the other, which is separated from Heiðarbaejarheiði by a whole series of mountains, one would need to have both outstanding stamina and outstandingly bad luck indeed. The remote place names mentioned so far are located far north of Heiðarbaejarheiði. However, the range of the 'Loss of Men' narrative also extends far to the south: there are at least two testimonies that connect the tragedy on Heiðarbaejarheiði with the naming of Þ orarinsdalur, which at its closest point lies some six kilometres to the south-east of Heiðarbaejarheiði and is separated from it by deep valleys and high ridges. The older of these accounts is found in an undated manuscript by J ohann Hjaltason (1899-1992) that gives an account of the farm of Fell: 16 . . . Þ orarinsdal [ur] . . . Dalurinn er kenndur við einn hinna 18 manna, sem maelt er að eitt sinn hafi orðið uti a Heiðarbaejarheiði, a leið til vers vestur að Isafjarðardj upi. . . . Þ orarinsdalur ('Þ orarin's Valley') . . . The valley is named from one of the eighteen men about whom it is said that they once died of exposure on Heiðarbaejarheiði, on the way westwards to Isafjarðardj up to a fishing station.
Geographically speaking, here the story has got completely lost, even more than was already the case in Filippus's narrative. Normally, the narrative states that the men who died were travelling south-west from Steingr ımsfj€ orður, to the fishing grounds of Breiðafj€ orður and specifically those around Snaefellsj€ okull. Here, the destination suddenly is Isafjarðardj up in the far north-west; the story has been turned around by ninety degrees. This has the consequence that the men not only get lost on the road, but that the road over the mountain Heiðarbaejarheiði leads in the wrong direction to start with. Of course, that something had gone wrong here did not escape attention. Bj€ orn is well aware that the story has become overstretched, and voices his resulting doubts. The story of the 'Loss of Men' has turned out to be so flexible that it has engendered place-name explanations that were intrinsically implausible even to those who told the tales. I have already quoted a substantial number of attestations of the 'Loss of Men' narrative, and even more could be added. The Ism us database of Icelandic cultural heritage contains a sound recording of an interview from 1970, in which Guðr un Finnbogad ottir (1885-1972), who lived on a farm in the area, gives another example of an identification of the victims of the accident as schoolboys. 18  This testimony is not without its challenges. Stef an Dan ıelsson refers to a published account in the regional annual journal Strandap osturinn (The Strandir Post), which, like Guðni's account, illustrates the tight interweaving of written and oral tradition. 20 More importantly, his account illustrates that the tradition of thinking about local place names in terms of 'Loss of Men' continues to this day, adding testimony after testimony and place name after place name-for Stef an gave his statement orally, and the field-name Eir ıksv€ ollur that Stef an mentions does not appear to be attested before the 1980s. 21 In fact, when I myself was making enquiries about the whereabouts of place-lore sites in 2019, it was pointed out to me by J on J onsson of Kirkjub ol that Hj alparsteinn, a big boulder next to an old road on the land of Tr€ ollatunga, also is connected to 'Loss of Men' (Figure 5). Hj alparsteinn ('Stone of Help') is said to be so-called because a sole survivor of 'Loss of Men' was found there, half-dead, and then received help. While several surveys of the place names of Tr€ ollatunga exist, none has mentioned this tradition before. 22 Given this continuing productivity of the tradition, any list of toponyms connected with 'Loss of Men' can only be preliminary.
'Loss of Men on Heiðarbaejarheiði' as a Historical Account?
'Loss of Men' is one of the 'great' stories of Steingr ımsfj€ orður. 23 It has grown deep roots not just on the highest mountain of the fjord's south coast, but also in a score of smaller places in a radius of more than ten kilometres around this mountain. The manner in which these places are connected with the story is very stable: in almost all cases-the exceptions being the 'Stone of Help' (Hj alparsteinn) and the 'Dead Man's Fields' (Dauðsmannsfoldir)-a place name formed with a male personal name is explained by the assertion that at this place the body of the man in question was found, who is identified as one of eighteen men who tried to cross Heiðarbaejarheiði. Pretty much everything else can be in a state of flux: the destination may be Snaefellsnes or Isafjarðardj up, the victims of the tragedy may be fishermen or schoolboys. In some cases, even the mountain where the tragedy occurred can fluctuate, as at least two attestations locate the 'Loss of Men' on Tr€ ollatunguheiði. 24 This fluctuation-or rather, what fluctuates and what does not-throws an interesting spotlight on the Sitz im Leben (place in life) 25 of the story. Almost the only stable element of the story appears to be that it is based on a reflection about place names, and exactly this seems to be its main Sitz im Leben: the vast majority of the attestations of the story that I have encountered are short explanations of one or two names by allusion to a tragedy alleged to have occurred on Heiðarbaejarheiði. 26 In the centre of these tellings stand the individual place and its name, which is thought about and then speculatively connected to a disastrous attempt at crossing the highest mountain of the fjord.
At this point it is worth addressing how realistic the narrative is that links these brief statements. Are the occurrences described in it historical? An important clue is provided by practicalities. From a practical perspective, it was indeed advantageous to cross the mountains by following a ridgeline, as snow did not accumulate there as much as on the valley bottoms and as on the ridgeline one avoided the risk of avalanches, which were greatly feared (J on J onsson, pers. comm., 2019). This is the reason why the routes marked on the maps created by the Danish General Staff between 1905 and 1915 so often follow the high ground, 27 whereas modern roads, made for automobiles, prefer the low ground. However, even to cater for this legitimate practical reason, travelling over Heiðarbaejarheiði would be overkill: to achieve the desired effect of avoiding deep snowdrifts and avalanches, the trail needs to follow the high ground, but it does not need to follow the ridge of the highest mountain in the area. The next ridge that would serve the purpose is Tr€ ollatunguheiði, which is only two kilometres from Heiðarbaejarheiði and has all the same advantages minus its disadvantages. Being almost 150 metres lower, as well as much gentler in its ascent, Tr€ ollatunguheiði is much easier going than Heiðarbaejarheiði. It also runs in exactly the same direction. As far back as records go, Tr€ ollatunguheiði has always been the route of choice for crossing the massif; the row of massive cairns that mark the highland trail to this day-now running parallel to a road and to the power lines of the Icelandic gridwas already marked on the map of the Danish General Staff drawn in 1912. The supposed road over Heiðarbaejarheiði, in contrast, has never been marked with cairns. There are a few cairns on the mountain, but they do not align to mark a route, and comparison with the map of the Danish General Staff shows that by far most of them have been erected by Danish land surveyors to mark triangulation points-most prominently in the case of the massive summit cairn that was built to mark the six hundred-metre point and which is widely visible from the mountains surrounding Heiðarbaejarheiði ( Figure 6). 28 So there is neither documentary nor material evidence that Heiðarbaejarheiði was ever used as a route to cross the mountain range. All mentions of Heiðarbaejarheiði as a route to Breiðafj€ orður that I have been able to find speak of a past use of the mountain as a crossing place; there appears to be no actual contemporary testimony that the mountain was ever used as such.
Interestingly, what little purported evidence there is for a former use of Heiðarbaejarheiði as a route to cross the mountains is itself tied up in a place name. In a description of the farm of Miðdalsgr€ of at the foot of Heiðarbaejarheiði, one finds the following statement: 29 Leiðar€ oxl dregur nafn sitt af pv ı, að par var farið upp a Heiðarbaejarheiði, sem aður var alfaraleið vestur ı Geiradal.
Leiðar€ oxl ('Shoulder of the Way') gets its name from that, that one went up onto Heiðarbaejarheiði there, which used to be the generally used way to the west into Geiradalur.
In a manner quite similar to the derivation of toponyms formed with personal names of dead men, this statement correlates the place name Leiðar€ oxl ('Shoulder of the Way') with the use of Heiðarbaejarheiði as a through-route. Thus, the place name is turned into evidence for a now-defunct road over the mountain to Breiðafj€ orður. Yet while at first glance this sounds all very logical, it is not the only way to interpret the place name. Leiðar€ oxl is a section of the slope on the northern end of Heiðarbaejarheiði, and its location towards the end of Heiðarbaejarheiði, just below the point where the slopes of the mountain turn into perpendicular cliffs, suggests that it may well have been a route over the mountain, but perhaps not quite in the way in which the previous statement imagines it. Rather, Hafd ıs Sturlaugsd ottir of the Environmental Institute of the Westfjords assumes that Leiðar€ oxl is the point where an old route, coming from Miðdalur, crossed Heiðarbaejarheiði in order to reach the trail over Tr€ ollatunguheiði some two kilometres further on. Hafd ıs has indeed been able to find traces of an old bridle-path which crosses Heiðarbaejarheiði at Leiðar€ oxl (pers. comm., 2019), and a journal article from the late 1980s mentions Leiðar€ oxl as the point where a path crossed Heiðarbaejarheiði that led from the now-abandoned farm of Tindur in Miðdalur to Tr€ ollatunga (G ısli J onatansson 1989, 124). Thus, it seems that the place name Leiðar€ oxl indeed reflects a path that led over Heiðarbaejarheiði; but this was not a path that went length-wise over Heiðarbaejarheiði to Breiðafj€ orður, merely one which crossed it at a comparatively low and easy spot in order to access the easier road over Tr€ ollatunguheiði.
Furthermore, not only is there no evidence that Heiðarbaejarheiði, against all considerations of practicality, was ever used as a way to reach Breiðafj€ orður, but the story also displays common elements of traditional folk narratives of the more openly fictional kind. One of these is numerical. In 'Loss of Men', it is a group of eighteen fishermen that is lost. This number is a very typical feature of Icelandic folktales, where companies of men normally consist of either eighteen or twelve members (Einar Olafur Sveinsson 2003, 293). In parts of the tradition, the men turn into schoolboys, reflecting a topos of Icelandic storytelling in which the trip to school is a recurring reason for long-distance travel (for instance, J on Arnason 1961,  At the same time, the account is also characterized by an element tragically rooted in real-life experience: death from exposure in north-west Iceland was a sadly common occurrence. To pick just a few examples from the region from the files of the Place Name Institute, there is a tradition recorded in 1953 according to which on one occasion the farmer owning the farm of H olar froze to death on his own land. 30 In the summer of 1911, the body of a young woman was found on the land of the farm Hr ofberg; she had frozen to death in the autumn of the previous year, when she was overtaken by a snowstorm. 31 On the land of the farm of Kleppustaðir, there is a cairn called Bj€ ornsvarða ('Bj€ orn's Cairn'). It got its name when the body of a certain Bj€ orn was found there. He was one of two men who froze to death on Maundy Thursday in 1865; the body of his companion was never recovered. 32 In the valley of Sel ardalur, the meadow Guðfinnuengi ('Guðfinna's Meadow') is said to be named from a woman called Guðfinna who died there of exposure. 33 Kattardalur ('Valley of the Cat') was re-named Ragnarsdalur ('Ragnar's Valley') when a certain Ragnar froze to death there around the year 1906. 34 The farm Kolbjarnarstaðir is said to have been abandoned after the last farmer, together with almost all members of his household, died when they were overtaken by a snowstorm while crossing the mountain Staðarfjall on the way home from a Sunday service in the church at Staður. 35 On 7 December 1925, a sixteen-year-old boy died of exposure in Brunngilsdalur, only a few hundred metres from the buildings of the next farm, when, after a particularly warm autumn, the livestock was gathered later than normal and the herding party was overtaken by a snowstorm. On the morning of that day, three went out to drive home the sheep, but only two came back. 36 One could continue this litany of death, but the point is clear: death from exposure, especially in the mountains, was a very real danger. Viewing 'Loss of Men' against this background, it is possible to pull the various strands of the preceding discussion together into an overall interpretation of this story.

An Imaginary Road and the Journey through Fear
In 'Loss of Men', we are not facing a simple account of an occurrence 'as it really happened'. Heiðarbaejarheiði, in all likelihood, never was used as a through-route to the south; the mountain would have been too difficult for that, especially with a functionally equivalent and much easier alternative available at nearby Tr€ ollatunguheiði. Yet, as the highest mountain of the area, dominating its surroundings, Heiðarbaejarheiði attracts attention and inspires the imagination. This imagination, being that of people deeply familiar with the local mountains and their weather, would immediately see how exposed one would be up on the ridge. Guðni J onsson's account explicitly muses on this point, stating: 'the weather up there must be quite extreme' ('allst orviðrasamt hl ytur að vera uppi par'). In Strandir, however, the idea of being exposed to the weather not only stands for discomfort, but for the real danger of death by exposure, a risk which was constantly present and again and again demanded its victims. This danger would have been a very real one on Heiðarbaejarheiði, and thus a story that had this danger at its core could easily latch onto the towering mountain, or maybe, rather, the towering presence of this mountain could suggest a way of putting this fear into words. At the same time, as soon as the idea of death on Heiðarbaejarheiði, as a formulation of the collective fear of death in the mountains, would have taken root, a multitude of toponyms derived from male personal names could suggest the identities of the victims. This combination of factorsthe reality of the threat of death by exposure, the dominating presence of the mountain, and the presence of toponyms derived from personal names-thus seems to have combined to form one big story about a tragedy that reflected the extent of the fear of death by exposure and befitted the size of the mountain to which it became attached. Narratively, the biggest mountain became the place of the biggest tragedy, and in this way it could tell the story of one of the region's great fears.
I suppose that the story owes much of its success to the specific local combination of factors: the physical dominance of Heiðarbaejarheiði, the real-world death toll taken by exposure, and the presence of place names suggestive of the names of possible victims. These factors made the narrative plausible and allowed it to act like a vortex into which ever more place names were drawn. Having this effect, 'Loss of Men' could even be described with Albert Esker€ od's concept of a 'tradition dominant': an element which becomes so dominant that it replaces other elements in the storytelling tradition, as when the mermaid on the Halland coast of Sweden becomes the standard supernatural entity that stories are told about, while elsewhere the same stories are told about different entities (Esker€ od 1947, 79-81; Honko 1981, 23-24;Tangherlini 1990, 378). The case of 'Loss of Men' could be described as a legend that operates as a tradition dominant for the interpretation of place names. And in the case of this particular tradition dominant, we can observe how a genuine collective fear crystallizes into a narrative told through place names interpreted as the names of victims.
It should be noted that, of course, we do not know what the 'real' historical basis of the various toponyms is. Þ orarinsdalur, P alshvammur, Halld orshvammur, and Þorsteinssund are too far from Heiðarbaejarheiði to be named from men who died while trying to cross it; and if these place names had other origins and later were reinterpreted, the same may be true for other toponyms as well. Indeed, the toponymy of Strandir contains more place names said to be derived from victims of 'Loss of Men' than 'Loss of Men' is said to have had victims. Even just the names discussed in this article would reflect more than twenty victims, and further enquiry might turn up more such toponyms. The prime mover, or at least the strongest moving force of the narrative, appears to be the place names in the real landscape: where a toponym is derived from a male personal name, it can respond to the magnetism of Heiðarbaejarheiði and can be read as a toponymic epitaph to a victim of the mountain. A story constructed out of collective fears thus is activated by place names, and in this way turned into local history. In current theorizing, place names are often conceptualized as 'mnemonic pegs' serving historical memory (for example, Brink 2013, 36). These place names, however, do not commemorate the storybecause it never 'really happened', even though tragedies like it happened all too often-but they form the catalyst through which it gets presence and reality.
In his ground-breaking survey of Icelandic folk narrative, The Folk-Stories of Iceland, Einar Olafur Sveinsson noted that '[p]lace-name stories come to life when the placenames come under discussion or when one sees the places' (Einar Olafur Sveinsson 2003, 67). This tallies both with my own experience in Strandir and with the pattern observed again and again in the material presented in this article: the vast majority of the attestations of 'Loss of Men' have the form of very brief summaries used to explain one or two local place names. Thus, 'Loss of Men' provides a remarkably well-documented example of how such storytelling works, and how a living oral tradition to this day keeps working alongside the published written versions. A multitude of 'small' local stories combine to form a web of storytelling until the local stories cumulatively create a story of regional importance. In the individual placefocused formulations of the story, we can observe a huge extent of variation as common local storytelling motifs are exchanged with each other, but we can also observe elements that remain constant throughout this variation: the close connection established to the environment through place names and the impact of the collective fear of death on the mountain. In this way, undertaking an ethnography of an imaginary road like the one over Heiðarbaejarheiði not only teaches us about the mechanics of such legends, but also opens up a path to approach an aspect of the emotional life encapsulated in this storytelling tradition.