Journal of the Geological Society, London, Vol. 156, 1999, pp. 1045–1050. Printed in Great Britain. Discussion on transpression and transtension zones Geological Society, London, Special Publications No. 135, 1998, 1–14 S. Lin, D. Jiang & P. F. Williams write: Transpression and transtension are common geological processes that have been the focus of many recent studies. The international confer- ence on ‘Continental Transpressional and Transtensional Tectonics’ held in London, England in March 1997 and the Geological Society, London, Special Publication No. 135 with the same title, arising from the conference, are both very timely. We commend the conveners and editors (R. E. Holdsworth, J. F. Dewey & R. A. Strachan) for the success of both the conference and the Special Publication. We presented a paper at the conference, which is included in the Special Publication (Lin et al. 1998), discussing oblique transpression/ transtension (in the sense that the relative displacement across the zone has a dip-slip component) and the resultant general triclinic kinematics within such zone. The Special Publication contains two additional papers not presented at the conference (Dewey et al. 1998; Jones & Holdsworth 1998), which include discussion of triclinic kinematics. We welcome these additions since they provide an opportunity to discuss further triclinic kinematics, which we believe is probably common in nature and important in understanding fabric development (Lin et al. 1998; Jiang & Williams 1998). The paper by Dewey et al. (1998) is concerned with many fundamental aspects of transpression and transtension that we consider to be important for future studies. The paper includes a summary of monoclinic transpression/transtension theory and some statements derived from modelling of triclinic transpression/transtension. Unfortunately, it is not always clear which statements apply to monoclinic transpression/ transtension only and which apply to more general cases (including triclinic transpression/transtension). This has the potential to cause confusion, especially amongst non- specialists. We feel that it is important to clarify this point. Our discussion will emphasize the similarities and differences between monoclinic and triclinic transpression/transtension. We start with the definition of transpression/transtension. Definition of transpression/transtension. Dewey et al. (1998, p. 2) define transpression and transtension as ‘strike-slip deformations that deviate from simple shear because of a component of, respectively, shortening or extension orthogo- nal to the deformation zone’. This is an inadequate definition, because it does not include oblique transpression/transtension in which a component of dip-slip is involved. It is also inconsistent with their own usage of the term in their paper. As pointed out by Robin & Cruden (1994) and Lin et al. (1998), the boundary-parallel simple shear component in transpression/transtension zones can be horizontal (trans- current transpression/transtention) or oblique (oblique transpression/transtension) (Fig. 1). Oblique-slip is commonly observed in present plate boundary regions (e.g. McCaffrey 1992), and strain geometry in many ancient shear zones suggests possible oblique relative slip at the boundaries. If the biaxial stretching axes of the zone boundaries ( ˙ a and  ˙ c in Fig. 1) are either horizontal or down-dip, as shown in Fig. 1, transcurrent transpression/transtention will have monoclinic symmetry at the bulk scale because the vorticity vector is parallel to one of the stretching axes ( ˙ c in Fig. 1c). In contrast, oblique transpression/transtension will generally result in de- formations of triclinic symmetry because the vorticity vector is oblique to the stretching axes (Fig. 1b). The only exception to the latter is when the boundaries of the zone are being stretched equally in all directions (the strain ellipse on the boundary is an expanding/shrinking circle, i.e.:  ˙ a =  ˙ c in Fig. 1b). In this case, the deformation is monoclinic irrespective of the shear orientation (Jiang & Williams 1998). If the stretching axes are neither horizontal nor down-dip, both transcurrent and oblique transpression/transtension generally lead to tri- clinic deformation and monoclinic deformation is a very special case that occurs when the vorticity vector happens to be parallel to one of the bixial stretching axes or when  ˙ a =  ˙ c . Therefore, we believe that transpression/transtension commonly leads to triclinic deformation and monoclinic deformation is expected to be a special end member. Strain pattern. Dewey et al. (1998) only summarize the strain pattern for monoclinic transpression/transtension zones. Taking this further, the strain patterns of monoclinic and triclinic transpression/transtension zones have similarities and differences (Lin et al. 1998). In the case of isochoric, constant strike length deformation, transpression produces oblate strain (k<1) and transtension prolate strain (k>1), no matter whether it is triclinic or monoclinic (Sanderson & Marchini 1984 and Fossen & Tikoff 1993 for monoclinic; Lin et al. 1998 for triclinic). Where the strike length is allowed to change, transpression may produce prolate strain, but only when the strike length is shortened, and transtension may produce oblate strain, but only when the strike length is stretched. For a general transpression/transtension zone of either monoclinic or triclinic symmetry, with or without strike length change, the following statement is true. Transpression produces oblate strain and transtension produces prolate strain, except when the strike line and the dip line have opposite signs of longi- tudinal stretching (i.e. in Fig. 1  ˙ a ·  ˙ c <0). This can be easily verified for general cases by using the unified position gradient tensor presented in Jiang & Williams (1998) and verified for monoclinic cases by Fossen & Tikoff (1998). Dewey et al. (1998) emphasize the bouncing of finite strain paths off k =0 and k =  axes of the Flinn plot and the coinciding ‘switching’ or ‘swapping’ of the finite strain axes between horizontal and down-dip in isochoric, constant strike length monoclinic transpression/transtension zones with angles of convergence () less than 20 (or kinematic vorticity number W k >0.81, or the simple shear/pure shear ratio ( ˙/ ˙ b ratio in Fig. 1)>2.76). This ‘bouncing’ and ‘switching’ does not occur in triclinic flow. In the latter, the finite strain paths plot away from k =0 and k =  axes (see fig. 10 of Lin et al. 1998). With increasing transpression/transtension obliquity ( in Fig. 1b or the angle between the direction of the zone boundary-parallel shearing and the strike of the zone), the paths plot closer to the 1045