Constructing the ‘legalities’ of encroachments in dam, canal, and stream reservations in the north-central province of Sri Lanka

ABSTRACT Encroachers on dam, canal and stream reservations in Sri Lanka have a long history, although largely neglected and sparsely investigated by academics and policy makers alike. They have been observed to degrade water resources at an alarming rate and have proven resistant to attempts of eviction. The present work attempts to employ a discourse analytical viewpoint to open up and explore the ideas, perspectives and justifications held by various social groups regarding the legality of encroachers, thus moving beyond the simplistic legal/illegal binary that has characterized previous academic inquiries on the matter. In doing so, the present work has found three competing discourses or ‘storylines’, each of which characterize encroachers in a different way. Together, these different discourses provide a platform to better understand the social practices, behaviors and motivations of the encroachers, and provide more complex and nuanced answers to the questions of ‘who they are’ and ‘what can be done about them’.


Introduction
Encroachers on dam, canal and stream reservations in Sri Lanka have a long history, although largely neglected and sparsely investigated by academics and policy makers alike. That encroaching happens in stream reservations has been noted by Chandrasekara and Gunawardena (2011); and further explored in relation to other forms of hydraulic infrastructure (dams and canal-networks) by Paranage (2017). While accurate numbers denoting the prevalence and density of illegal encroaching on reservation areas have not been recorded; preliminary statistics from the Divisional Secretariat of Sri Lanka along with information from the Ministry of Environment and Natural Resources, Sri Lanka, enables us to piece together a picture that presents encroachers as an active and significant threat to the hydraulic apparatus supporting the irrigated agriculture of Sri Lanka (Peebles 1990;Mendis 1999;Nelson 2003;Zubair 2005;Lanka 2014). This is underscored by the fact that encroachers are most often found in the north-central province of Sri Lanka (see Figures 1 and 2 for a representation), which has been the prime focus of all state irrigation development activities throughout recorded history (Leach 1959;Bandara 2000). Encroachers on dam, canal, and stream reservations have been observed to degrade water resources at an 'alarming rate' (Chandrasekara and Gunawardena 2011, 135) by (a) clearing the forest of reservation areas, rendering the canals, dams and streams vulnerable to soil erosion and stream bank erosion, (b) dumping solid waste into the water resources, thereby polluting water quality.
As a result of the threat they pose to the hydraulic infrastructure; illegal encroaching on canals, dams and streams have been stringently categorized as 'unlawful', a categorization effected on the basis of many laws and regulations governing irrigated agriculture and water resource management. Given the considerable importance attached to preserving irrigated agriculture and its supporting hydraulic network, 1 successive administrations of Sri Lanka have enacted numerous laws and regulatory instruments to protect both agriculture and water resources: from the Irrigation Ordinance No: 32 of 1856 (which was the first enactment developed by the British colonial administration to legalize traditional irrigation and water management practices), more than 50 statutes and circulars have been decreed in relation to the development and management of water resources along with the establishment of a large number of enforcement institutions (Samad 2005). The cumulative effect of such laws and regulatory instruments has been to characterize the identity and social practices of those who encroachboth 'who they are' and 'what they do'as being borderline or outright illegal.
While not denying the 'illegality' of encroachers per se; nor denying the detrimental effect that they have on water resources of Sri Lanka; the present study invites looking at the ostensive 'problem' of encroachers from a different perspectiveusing the theoretical insights of critical discourse analysis. Instead of considering encroachers as 'illegal' based on the legal/illegal binary put forward by traditional institutionalist perspectives, this study looks at encroachers, and more specifically the relationship between encroachers and the legal system that defines them as such, when situated amidst competing 'discourses' and discursive fields. Using critical discourse analysis in this context may prove useful in recognizing the underlying complexity of encroachers as a social group whose identity and actions should be constructed in relation to the legal processes; not through them. At a more practical level, critical discourse analysis may also be of use in understanding why encroaching continues to be a persistent problem despite the wide array of legal and regulatory instruments and enforcement institutions established to impede their growth.

Theoretical framework
Discourses are broad and encompassing theoretical instruments. Hajer (1995, 44) defines discourses as 'ideas, concepts and categorizations that are produced, reproduced and transformed in a particular set of practices and through which meaning is given to physical and social realities'. In other words, discourses are essentially political (Gee 1999) representations of things in the material world. We need discourses to make sense of thingsto construct reality. At a very fundamental level; discourses govern the way we provide meaning to aspects of the material worldwe inevitably endorse a particular perspective as to what the 'world' is likeoften embedded in our use of language (Gee 1999(Gee , 2004(Gee , 2014 and from within this perspective (i.e. 'discourse') we define or provide meaning to the things in the world around us. In short, while aspects of the material world may exist outside of discourses; we cannot understand them from outside our own perspective(s) or discourses. This is an assertion that acts as a theoretical nucleus particularly for discourse analysis, and more generally for the episteme of social constructionism and phenomenology (Burr 2015;Miller 2017).
A second, related, assertion that is useful for the theoretical framework of the present work is the political consequences of recognizing the multiplicity of existing discoursesor the multiple frames of references. It is not sufficient to say that discourses create or structure 'reality' and imbue it with meaning; it is equally important to realize that different discourses represent reality in vastly different ways; or more specifically, that different discourses provide very different meanings to the same set of objects, circumstances, persons and activities (Wodak and Meyer 2009). This is important theoretically since it displaces the single question single answer models in social and political inquiry (Hajer 1995;Harvey and Braun 1996); and opens up hitherto closed aspects of the world. Indeed, discourses may interact, overlap, merge or displace each other (Burman and Parker 1993), or might be engaged in a constant struggle for discursive hegemony (Boelens and Vos 2012). Relating to the present study, the multiple-discourse standpoint enables us to open up many possible ways in which the 'problem' of encroaching can be understood; without being constrained by pre-determined 'limits' as to what can be said on the matter. Most importantly, it enables us to look past the legal/illegal binary that characterized encroachers and drives us instead to look at the complex construction of the 'legalities'how different groups of actors, interests and actor-politics may provide different meanings to the laws that govern, limit and exclude encroachers. This line of inquiry regards both the laws and encroachments as 'contested spaces' to be discursively analysed.
Based on the above mentioned epistemological and theoretical premises, what the present study sets out to do is to identify the dominant discourses that exist with regard to encroachers on dam, canal and stream reservations in the north-central province of Sri Lankabe it 'inclusive' or 'exclusionary' (see Saff 2001) in terms of seeking to either accept/include or reject/exclude encroachers as legitimate stakeholdersand identify the political implications arising from subscribing to these discourses. Specifically, I ask the following questions that will guide the methodology, findings and discussion of the present work: (1) How does each discourse characterize or define encroachers, the laws enacted in relation to encroachers, and the purported 'problems' that they pose? (2) Who are the actors that subscribe to these discourses and why? (3) What are the political and social (including behavioral) implications that arise from these discourses and how are they significant?

Materials and methods
The present work takes, as the study sample, two villages located in the north-central province of Sri Lanka. There are three main reasons for this selection. Firstly; the north-central province is considered to be the heart of Sri Lanka's irrigated-agriculture network (and has been considered in this way, for several centuries) that houses the centrepieces of hydraulic infrastructure (De Silva 1981) including some of the most notable dams and reservoirs (Azmi 2007;Chandrasekara and Gunawardena 2011). Secondly, the north-central province of Sri Lanka has been the subject of large scale water engineering projects (for example the Mahaweli Development Programme, which is documented as the largest national development project ever to be undertaken in the country) that sought to improve the province's hydraulic network to maximize agricultural production (Mendis 1973;Azmi 2007). Thirdly, and as mentioned in the introduction, the north-central province accommodates the highest rates of encroachments recorded (Chandrasekara and Gunawardena 2011; Paranage 2017) relative to the other provinces of the country.
The north-central province contains two administrative districts (Anuradhapura and Polonnaruwa) and 21 villages (Wickramasinghe 2015). Both these districts contain significant hydraulic infrastructure and have been the focus of improvements by the Mahaweli Development Programme since 1970. Of these two districts, I have randomly selected two villages (Thambuttegama and Jayanthipura) one representing each districtsee Figures 1 and 2 for their locations. I collected ethnographic data from three different groups of people involved with (or dependent upon) the hydraulic network at various levels: (a) from administrative officers of the Mahaweli Development Programme who are involved with the maintenance and upkeep of the hydraulic infrastructure and to further agricultural improvements in the project area; (b) from farmers who are directly dependent on the hydraulic infrastructure and water network to sustain their livelihoods; and (c) from encroaching families who can be considered as 'outlier' groups in the sense that they, much like the farmers, are dependent on the water network to sustain their livelihoods, but are not part of the 'official' hydro-irrigation network (in fact, they are legally excluded and prevented from accessing the network).
The identification of the three groups of stakeholders (involved with the hydraulic network at various levels for various purposes) was done intentionally, and with reference to politics of scale (Herod and Wright 2008) so as to maximize the opportunity to capture distinctive discourses. However, it should be mentioned that there was no attempt to be exhaustive in attempting to capture either discourses or stakeholders. In fact, Hajer (1995) and Gee (1999) have demonstrated that most discourses are relatively open ended without clear boundaries for easy demarcation. As such, the intention of the present work is to identify and contrast several perceivable and large scale 'discourses'or to borrow a term from Hajer (1995): 'story-lines'that can be qualitatively contrasted with each other. I allow for the possibility that the discourses identified in the analysis may be composed of smaller sub-discourses and that a deeper analysis may discover additional layers of discourse interaction for analytical purposes. This possibility, however, does not detract from the object of the present work; which is to provide a starting point for understanding and theorizing about the underlying complexity of encroachers as a 'social group' whose identity and actions should be constructed in relation to the legal processes rather than on the basis of laws, regulations and institutionalized norms.
Finally, the tools used to collect data were a combination of semi-structured interviews (N = 10*3); Focus Group Discussions (N = 2*3) and participant observation in relation to each of the three groups. A field diary was used to record observations. Further, since encroachers are a group that is 'labeled' in relation to the existing laws and regulations that govern irrigation systems and water-networks, I have conducted a documentary analysis of a number of statutes and regulations including the Crown Land Ordinance of 1846; the Land Development Ordinance of 1935; the Waste Land Ordinance of 1973; the Irrigation Ordinance of 1949; and the National Aquatic Resources Act of 1981.

Results
The findings revealed that there were three main discourses present in the research field. The salient features of these three discursive fields are presented below, in Table 1, organized with reference to the research questions: (1) how each discourse characterize encroachers, the laws enacted in relation to encroachers, and the problems that they pose (especially how these factors are represented through language); (2) what actors subscribe to these discourses; and (3) the political and social implications that arise from these discourses.

Discourse 1: exclusion of encroachers based on economic and ecological justifications
One of the recurring themes used to justify the exclusion of encroachers has been to highlight their lack of concern for the natural environment and the hydraulic network that underlie the irrigated agriculture. This discourse was dominant among the officers involved with the hydraulic network in an administrative capacity, especially the officers of the Mahaweli Authority of Sri Lanka, tasked with maintaining and running the irrigation system and ensuring that water is efficiently distributed through the hydraulic network of reservoirs and canals. Much of the officers' opposition to the encroachers were couched in terms of an economic and ecological discourse. The exclusionary attitude based on ecological premises is reflected in the following comment of one officer interviewed: … [These] encroaching families … produce a substantial amount of sewage that will be dumped directly into the canals that distribute water [and] create a ripple effect on the whole ecosystem … . This is going to leave us with an irreparable water network, and cause diseases and mosquito infestations. This [small] group of people might eventually destroy the sensitive ecological area … Field notes (2017), translated In addition to these sentiments, there was also the fear that encroachers, by clearing spaces reserved for the protection of the hydraulic apparatus (such as the forest and shrubs that cover the perimeter of the reservoirs to stop soil erosion) might also cause the eventual destruction of the hydraulic infrastructure and degradation of water resources.
In parallel with the ecological discourse (and often connected to it) runs the economic discourse that exclude encroachers based on economic justification. To quote an officer: … The encroachers are not represented in the official map of settlers. However, they continue not only to create housing structures, but also to cultivate. The water that we release into the farms through the canal network each day is precisely calculated to match the needs of an [X] number of families. We can't change how much water we release into the system; those amounts are determined at the very beginning of the [bi-annual] season by the irrigation ministry of the central government. What encroachments do, is that they create [illegal] water inlets into the official canals and siphon off the official supply. When that happens; there isn't enough water to go around to everyone … Field notes (2017), translated As can be seen, both ecological and economic concerns share some common themes. Broadly speaking, the administrative officers tend to characterize water as a resource which is limited. Critical discourse analysis is particularly helpful in providing an explanatory structure for this 'perception of water as a (scarce) resource': the officers are representatives of the Mahaweli Project, and the objective of this project is to maximize irrigation production and development for Sri Lanka. Being part of the Mahaweli project, the officers can be seen as thinking within the project practices and internalizing the goals or values of the project. In an earlier section, we defined discourses in terms of 'ideas, concepts and categorizations that are produced … in a particular set of practices' (Hajer 1995) it is clear, then, in this instance why representatives of the Mahaweli project would define water in terms of a 'resource' and encroachers as a 'problem'. After all, encroachers are a perceived threat to the viability of the project objectives (agricultural development) in terms of both ecology (degradation of water quality) and economy (less efficient water distribution leading to stagnant cultivation). As such, encroachers are discursively constructed as a 'problem' that needs to be 'expelled' to ensure the 'greater good' of the people. This perception of encroachers as an economic and ecological problem translates into political action by communicating notices of eviction to the encroaching families, and on several occasions, following up on those notices by actually evicting them. These actions on the part of the officers are seen as justified on legal authority. However, what is interesting are the justifications given: … The laws are there to protect the people, and the environment. We have more than fifty laws that we use to evict encroachers … they are a problem to the environment and the economy. Besides, what they are doing is illegal; they are causing harm to the economy and other people … [We] sympathize with encroachments, but we also have to look at what will happen if they continue to stay and destroy the economic prospects of everyone … Field notes (2017), translated Here, what is interesting is not the statement of the obvious (that encroachers are rendered 'illegal' by the laws) but the officers' interpretation of the laws, and the way they justify enacting the laws with referring to economic/ecological concerns. The economic basis of the laws, of course, can be made clear in an examination of the various articles of legislation that deal with encroaching in some way (for a list of selected sources, see the methods section). Further, given that (a) irrigated agriculture comprise a significant aspect of Sri Lanka's economy; (b) Sri Lanka as a country is identified as a 'hydraulic civilization' and a country of economic-farmers; and (c) successive rulersthe pre-colonial era monarchy, the colonial administration as well as post-colonial governments all promoted irrigated agriculture as the principle means of economic development, it is no surprise that one can easily locate the raison d'être for anti-encroachment laws within an economic base (Uphoff and Wijayaratna 2000;Meinzen-Dick and Bakker 2001;Molden et al. 2010). Thus, the economy defines and justifies the laws against encroacherseven so, the officers appear to take the economic interpretation of laws to quite another level by drawing a parallel between economy and morality. There is a tendency to look at protecting the economy and environment in terms of 'greater good'; and encroachers as a form of 'evil' that needs to be expelled. Thus, encroachers are not simply illegal because the law dictates that they are illegal; they are illegal because allowing them to stay would be against the 'greater good' which is understood in terms of economy and ecology. In sum, encroachers are perceived an economic/ecological problem which is to be combated by laws which are there to ensure economic prosperity.

Discourse 2: exclusion of encroachers based on crime-related concerns
The second dominant discourse that was exclusionary towards encroachers was based on the idea that they increased crime rates in the neighborhood. Subscribers to this discourse were found mostly among the farming community in the two sample villages. The farming communities comprise several generations of farmers, who are part of the recognized water network and have ownership rights to their residential and cultivation land plots. This idea that encroachers were linked to crime rates was expressed by one farmer as: … I have had several break-ins in my house ever since the encroachments increased … there's a lot of petty pilfering, trespassing on private property and brazenly collecting water from private taps going on … destroying our water ways. There's a lot of drug peddlers with accompanying general disorders … and a risk of unrest.
There's always some of them walking around, looking for trouble … Field notes (2017), translated As this quote demonstrates, a predominant theme by farmers (landowners) is that encroachers are violating the sanctity of private property relations and are engaging in criminal activities within the village(s). However, the number of recorded incidents of crime in the area has not demonstrated a notable or acute increase, according to the police and other officers of law enforcement in these communities. It is premature to base a conclusion on this since a number of such crimes appear to go unreported, with no complaints being made. Farmers (landowners) do seem to agree, however, that encroaching constitute an 'illegal' phenomenon and need to be evicted from their neighborhood. Despite this general agreement, it is interesting to note exactly how farmers characterize or phrase this 'illegality': for them encroachers are illegal because they are associated with rising criminal activity. We noted above that encroachers in discourse 1 were characterized as illegal because they represented a growing economic and ecological concern; in discourse 2, however, they were associated with (often non-economic or ecological) 'crimes' such as trespassing and break-ins. In fact, very few (roughly 10 percent) of the farmers who were interviewed for the study even reflected on encroachments as an economic or ecological threatfor most of them, encroachers created a hazard for personal security, which served as the foundation for considering them as 'illegal'.
This attitude towards encroachers as a 'criminal element' was observed to create both overt and subtle conflicts between farmers and encroachers. Generally, there is an absence of connections or dealings between these two groups of people, despite living in relatively close proximity. Further, farmers have been known to act collectively as a pressure group to compel officials to take action against encroachers (i.e. carry out evictions).

Discourse 3: inclusion of encroachers based on equity
The third main discourse is found among the encroaching families themselves. This discourse is antithetical to discourse 1 and 2, in that it is inclusive towards encroachers based on equitability. In order to grasp and contextualize the arguments of encroaching families, a brief digression is necessary to outline the legal history of land attribution in the North-Central Province; especially in the Mahaweli Settlement Areas.
In studying the legal framework for the irrigation lands and water networks, I begin from the colonial period, where legislation for Sri Lanka was subsumed under the sovereignty of the British crown. The most important piece of legislation in this context is the Crown Land (Encroachment) Ordinance of 1846, which sets the basis for the regulation of all state lands including irrigated lands in the north-central province. This ordinance, first enacted in 1849, gave the crown the right to take over any uncultivated land and to grant, sell or lease such lands to individuals or institutions 'for any purpose which the Governor-General may approve'. As such, all forest, waste, unoccupied or uncultivated lands were presumed to be Crown property until the contrary was proved, and the land used for irrigated agriculture in the north-central province was no exception. This presumption in favor of the crown was very strong, so that very few peasants could win cases of land disputes against the crown. The legal situation created by this ordinance is of extreme importance since it identifies and delineates certain portions of land in Sri Lanka as 'crown land' or 'state lands' (Paranage 2017).
The second major development in the legal framework arrived in the form of the Land Development Ordinance of 1935 (hereinafter abbreviated to LDO 1935) which reflects the increasing concern for the development of irrigated agriculture in the north-central province, and for improving the position of the peasant cultivator. For the first time, the government sought to alienate crown lands to several classes of settlers, most notably the farmers settled in the north-central province. However, these lands are not given free of restriction; and in particular the lands so awarded were subject to an important rule (in accordance with section 35, 37, and 48A subsection 1 of LDO 1935): the rule of minimum subdivisions. This rule effectively places severe restrictions on farmers (landowners) who are cultivating in the north-central province both in terms of dividing land into smaller parcels and in leasing or selling their land to a third party.
Essentially, section 35, 37, and 48A subsection 1 of LDO 1935 (and the rule of minimum subdivisions) prevents farmers from fully 'owning' the crown lands. Section 35, in fact, stipulates that the (alienated) crown lands cannot be inherited by more than one person at any given time (this is a restriction in rem, meaning that the restriction against fragmentation runs with the land itself and binds any successive owner). Further, sections 37 and 48A stipulate that in the event of the original owner having more than one offspring, only one may inherit title to the land. The owner has the right to decide which offspring should inherit the land (he/she should specify this fact by lodging an application at the local land Kachcheriyathe district office dealing with land related matters) failing which the land title passes to the eldest child by default. While section 48, in theory, makes it possible for the landowner to petition the district secretariat (via a lengthy process of application) to separate his/her land among more than one offspring, the district secretariat holds discretionary authority over whether to concede to such a request (the petitions themselves are considered on a case-by-case basis so as to avoid creating a legally binding precedent). Whilst the present study has not attempted to empirically verify the rates of success or failure of such applications, it was anecdotally recorded that the lack of clarity regarding their legal position and the complexity of the application process acts as a deterrence to farmers ever attempting to divide up the land. We can surmise, that the above-mentioned provisions clearly come into conflict with the more usual practices of inheritance followed in other parts of Sri Lanka with reference to non-crown lands and has the net effect of rendering most of the second generation of settlers homeless, since (excepting some instances where the legally entitled offsprings allow the non-entitled sibling to stay with them in the family home) the non-entitled offsprings are usually left landless, to fend for themselves.
There are clear economic motivations for the rule of minimum subdivisions and the singular pattern of inheritance described above, since dividing up lands into smaller plots can cause land-fragmentation, thereby reducing cultivation output (see Teshome et al. 2016;Wadduwage et al. 2017;Xie and Lu 2017). We have already noted in the introduction and in relation to discourse 1, that in preparation of laws related to governing the land-water nexus in the north-central province, there has (historically) been a greater focus on irrigation development than on individual rights. However, from the perspective of individual rights to land, one can see that this dominant economic-centric discourse and the inheritance practices it manifests is very problematic; especially since it can be plausibly argued that the laws themselves appear to create the problem of encroaching: … These laws are manifestly unfair for us. My father was a farmer and landowner, but he never actually went to the land Kachcheriya [the district office where applications to alter title deeds needs to be lodged] to divide up this land between me and my two brothers. Now the officers say that I can't live in my own homewhere should I go, and why? Why should the land be my brother's right any more than it is mine? How is this just? Field notes (2017), translated Clearly, this quote has several important implications. Firstly, if encroachers are indeed part of the second or third generations of farmers (a 'lost generation'), then we can understand their position in considering the laws as unjust from an individual rights perspective or based on a 'natural law doctrine' (where certain laws are not regarded as 'real' laws if they contravene moral principles: see d'Entreves 2017). Indeed, the language use here (the words 'unjust'; and 'my right') does support the equitable discursive framework. Secondly, one can note that the equitability discourse does have important political ramifications in the sense that perceiving lands as a right that is owed to them may be the primary motivational factor that causes encroachers to stayeven braving repeated pressures of eviction. Lastly, it should be noted that both officers and farmers (while not disavowing the fact that at least some who encroach are part of the second or third generations of farmers) maintain that the majority of encroaching families are 'outsiders'. Even when they admit that a particular family of encroachers are relatives; farmers still tend to frame them as interlopers, or outcasts. This point has important theoretical implications that I shall refer to in the discussion section of the present work.

Discussion
My attempt in this study has been to open up the discursive complexities involved in constructing the 'legalities' of encroachments in dam, canal, and stream reservations in the north-central province of Sri Lankato understand the potential for multiple interpretations of the relationship existing between the law and encroachers and the political consequences such an understanding would entail. This section discusses the major themes that emerged from the present work and their implications.

Discourse analysis and the construction of 'legalities'
In the section on theoretical approaches, I discussed the epistemological position of social constructivism and phenomenology, and the contributions of discourse analysis in bringing to light the different concurrent representations of things in the material world. The findings in this work demonstrate an instance of how various discourses can represent the same set of laws, regulations and statutes in vastly different ways, and the political ramifications arising from these differences. Essentially, instead of perceiving laws and regulations banning the existence of encroachers instrumentally, the findings demonstrate that different discourses construct different 'legalities' based on various interpretations of said laws.
I have simplified the existing discourses into three main 'storylines' (Hajer 1995): (a) an economic-ecological understanding of the laws and regulations, (b) a crimebased interpretation of laws and (c) an individual rights-based perspective on laws. Each of these different discourses were perceived among different social groups (among officers of the Mahaweli project, farmers and encroaching families, respectively) who have their own social practices and exist within particular material and social frameworkslending credence to the definition of discourses subscribed to, as 'ideas, concepts and categorizations that are produced, reproduced and transformed within a particular set of practices' (Hajer 1995, 44, my italics). Further, each of these three storylines take the same set of objective laws and regulations regarding encroachers and positions them within their own perspective, so that the laws mean something different when situated within a particular perspective.
I have also been able to identify, via discourse analysis, certain techniques that are used to achieve this act (situating objective material facts within different perspectives to yield different meanings). For one, by analysing the language of the interviewees along with contextual information, we have noted that certain material facts are assigned more or less importance within each perspective. An example of this is how encroaching families ascribed standing to the fact that they are, in fact, the siblings and cousins of the farmers (landowners) in the area, while the farmers tended to discount this fact (unless directly confronted about it) and consider encroaching families as 'interlopers' or 'outsiders'. For another, I show that certain values (as opposed to facts) are likewise assigned different importance: both farmers (landowners) and officials of the project support the laws banishing encroachments, stating that they serve the greater (or collective) good of the community, while encroachments assert that the same laws are 'unjust' by attributing greater importance to a perspective on individual rights. Lastly, I show how (within these discourses) particular linkages are created between disparate elements: for instance, between ecology, economy and morality. Safeguarding the ecology or economy (through laws) is seen as 'good' (morally) within the first discourse, while disrupting the ecological or economic patterns is perceived as 'bad'.
The construction of different 'legalities' through discourses is also important from a political, action-based standpoint. We have seen that the various decisions made by actors of each social group (be it officers, farmers, or encroachers) can be understood with reference to the discourse(s) that they subscribe to. Particularly, in the case of encroachers, we have noted that their decision to stay is at least partially influenced by their desire to combat an unjust law. This finding goes against the conclusions of many observers who attribute the persistence of encroachers to 'deficiencies in implementing and enforcing laws' (Chandrasekara and Gunawardena 2011, 134). Hence, a case can be made for the value of discourse analysis in informing policy initiatives drafted in relation to encroachers, given (as demonstrated above) the power of discourse to institute, solidify, change, create and reproduce social formations (Wetherell and Potter 1992).

Discourse analysis and the politics of the 'other'
The findings of this work also contribute to the body of literature that puts forward the idea of 'otherness' and its centrality to sociological analyses of how majority and minority identities are constructed (see Said 1978;Pandey 2004;Anderson 2007;Marino 2015;Russell and Lyon 2015). In order for there to be 'otherness', there has to be a process of identity ascription by dominant groups (officials and famers) to the subordinate group (encroachers). Sibley (1992, 109) notes this as: … the perception of minority cultures as being beyond the boundary of 'society' is associated not only with characteristics of the group but also with particular images of particular places, the landscapes of exclusion which express the marginal status of the outsider group … Language becomes an important signifier in this process, allowing for the dominant group to ascribe various pathologies, such as propensities of criminality, to the other and in so doing, legitimate their exclusion from particular spaces in society (Saff 2001). In the present work, we have seen this in action, where encroachers are seen as predisposed to committing crime (Discourse 2) and creating 'diseases' and 'mosquito infestations' (Discourse 1). It is also significant to note that there appear to be no agreement between the two exclusionary discourses on whether encroachers bring about a downturn in economic prospects or increased criminality or both. Criminality was a state that was noted by the farmers; while the encroachments' propensity to create and spread diseases by dumping sewerage into canals was a fact noted by the official of the Mahaweli project. Thus, on the one hand, we can legitimately cast doubt as to the severity and degree of either of the charges (disease and criminality) laid on the encroachers while, on the other hand, we can agree about the fact that both Discourse 1 and Discourse 2 create analogous exclusionary models towards encroachers.
In the same vein, we can also argue that the words 'disease' and 'criminality' are employed in these discourses as metaphors, to justify the exclusion of encroachers as a social group. This point receives some support from Goldberg (1993) and Miles and Brown (2003) who argue that exclusion of social minorities is often justified by the dominant economic, political, ethnic, racial and cultural groups with reference to metaphors of pollution (e.g. impurity, dirt, disease, crime, etc.).

Concluding remarks
In the present work, I have explored the phenomenon of encroachers in dam, canal, and stream reservations in the north-central province of Sri Lanka and, by employing discourse analysis, moved beyond the legal/illegal binary to argue that laws are always interpreted based on a particular discourse; and different discoursesin essenceconstruct different 'legalities'. Different discourses provide different 'meanings' to the laws that characterize and exclude encroachments as well to the encroachers themselves. Thus, by opening up the number of possible ways to understand encroachers, one can begin to understand their behaviors, social practices and motivations in a more meaningful way. Further, I submit that discourse analysis, as evidenced in this case, is an asset in undertaking meaningful policies to address development related issues in general. Note 1. The importance given to protecting irrigated agriculture and the supporting water resource systems has been justified on two fronts. Firstly, irrigated agriculture accounts for 18% of the GDP and 26.4% of the workforce and is of considerable economic importance. Secondly, irrigated agriculture has had a history that exceeds 2500 years, with country's very identity being constructed as a 'hydraulic civilization' (Peebles 1990).

Disclosure statement
No potential conflict of interest was reported by the author.

Data availability statement
The datasets generated during and/or analysed during the current study are not publicly available due to sensitivity and ethical considerations but are available from the corresponding author on reasonable request.