With the increasing development of relations of consumption between discipline knowledge and students, educators face many pressures. One of these pressures is the emotional response of students to their learning experiences and the weight given to their evaluation of teaching by universities. This study emerged from the polarised nature of student responses to one particular area of study in landscape architecture, the integrative discourse of Landscape Planning. While some students found this subject highly rewarding, others found it highly confronting. Thus the main aims of this study are to describe how the students, teacher and institution construct this discourse and to propose a way to rethink these differences in student responses from a teacher's perspective. Firstly, the context of the study is outlined. The changing nature of higher education in Australian society frames the research problem of student-teacher struggles in Landscape Planning, a domain of knowledge in landscape architecture that is situated in a an enterprise university in Queensland. It describes some of the educational issues associated with Boyer's scholarship of integration, contemporary trans-disciplinary workplaces and legitimate knowledge chosen by the institution [Design], discipline [Landscape Architecture], teacher [Landscape Planning] and students [useful and relevant knowledge] as appropriate in a fourth year classroom setting. Secondly, the conceptual framework is described to establish the point of departure for the study. This study uses the work of Basil Bernstein, Harvey Sacks and Kenneth Burke to explore the changing nature of knowledge relations in Landscape Planning. Unconventionally perhaps, it begins by proposing a new concept called the 'decision space' formed from the conceptual spaces of multiple participants in an activity and developed from notions of creativity, conceptual boundaries and knowledge translation. It argues that it is in the 'decision space' that this inquiry is most likely to discover new knowledge about student-teacher struggles in Landscape Planning. It outlines an educational sociological view of the 'decision space' using Bernstein's concepts of the underlying pedagogic device, pedagogic discourse, pedagogic context, recontextualising field and most importantly the pedagogic code comprising two relative scales of classification and framing. It introduces an ethnomethodological view of knowledge boundaries that construct the 'decision space' using Sacks' concepts of context-boundedness and indexicality in people's talk. It also makes a link to a rhetorical view of knowledge choices in the 'decision space' using Burke's concepts of symbolic human action, motive and persuasion in people's speeches, art and texts. Thirdly, the study is divided methodologically into three parts: knowledge relations in official and curriculum texts, knowledge choices in student drawings and knowledge troubles in student talk. Knowledge relations in official texts are investigated using two relative scales of classification and framing for Landscape Planning and its adjacent pedagogic contexts including Advanced Construction and Practice 1 and 2 and Advanced Landscape Design 1 and 2. The official texts that described unit objectives and content in each context reveal that Landscape Planning is positioned in the landscape architecture course in Queensland as an intermediary discourse between the strongly classified and strongly framed discourse of Advanced Construction and Practice and the weakly classified and weakly framed discourse of Advanced Landscape Design. This seems to intensify the need for students in their professional year to access and adapt to new pedagogic rules, apparently not experienced previously. A further subjective reflection of my own week 1 unit information as curriculum text using classification and framing relations is included to explain what characterised the rationale, aim, objectives, teaching programme, assessment practice and assessment criteria in Landscape Planning. It suggests that the knowledge relations in my teaching practice mirror the weakly classified and strongly framed discourse of the official text for this unit, that is that students were expected to transcend knowledge boundaries but also be able to produce specific forms of communication in the unit. Knowledge choices in student drawings in Landscape Planning are described using a new sociological method of visual interpretation. It is comprised of four steps: (a) setting up a framing scale using the social semiotic approach of Kress and van Leeuwen (2005) (contact gaze, social distance, angle of viewpoint, modality, analytical structure and symbolic processes) combined with the pentadic approach of Burke (1969) (act, scene, agency, purpose); (b) setting up a classification scale using the concept of agent from the pentad of Burke (1969) combined with how the relationship between 'I' the producer and 'you' the viewer is constructed in each drawing, like a sequence in a conversation according to Sacks (1992a); (c) coding student drawings according to these two relative scales and (d) assessing any shifts along the scales from the start to the end of the semester. This approach shows that there is some potential in assessing student drawings as rhetorical 'texts' and identifying a range of student orientations to knowledge. The drawings are initially spread across the four philosophical orientations when students begin Landscape Planning and while some shift, others do not shift their orientation during the semester. By the end of the semester in 2003, eight out of ten student drawings were characterised by weak classification of knowledge boundaries and weak framing of the space for knowledge choices. In 2004, nine out of twenty-one drawings exhibited the same orientation by the end of the semester. Thus there is a changing pattern, complex though it may be, of student orientations to knowledge acquired through studying Landscape Planning prior to graduating as landscape architects. Knowledge troubles in student talk are identified using conversation markers in student utterances such as 'I don't know', 'I think', 'before' and 'now' and the categorisation of sequences of talk according to what is knowable and who knows about Landscape Planning. Student talk suggests that students have a diverse set of affective responses to Landscape Planning, with some students able to recognise the new rules of the pedagogic code but not able to produce appropriate texts as learning outcomes. This suggests a sense of discontinuity where students dispute what is expected of them in terms of transcending knowledge boundaries and what is to be produced in terms of specific forms of communication. The study went further to describe a language of legitimation of knowledge in Landscape Planning based on how students viewed its scope, scale, new concepts and other related contexts and who students viewed as influential in their selection of legitimate knowledge in Landscape Planning. It is the language of legitimation that constructs the 'decision space'. Thus in relation to the main aims of the study, I now know from unit texts that the knowledge relations in my curriculum design align closely with those of the official objectives and required content for Landscape Planning. I can see that this unit is uniquely positioned in terms of its hidden rules between landscape construction and landscape design. From student drawings, I acknowledge that students make a range of knowledge choices based on different philosophical orientations from a pragmatic to a mystical view of reality and that my curriculum design allows space for student choice and a shift in student orientations to knowledge. From student talk, I understand what students believe to be the points of contention in what to learn and who to learn from in Landscape Planning. These findings have led me to construct a new set of pedagogic code modalities to balance the diverse expectations of students and the contemporary requirements of institutions, disciplines and professions in the changing context of higher education. Further work is needed to test these ideas with other teachers as researchers in other pedagogic contexts.