Given any Euclidean ordered field, Q, and any 'reasonable' group, G, of (1+3)-dimensional spacetime symmetries, we show how to construct a model M G of kinematics for which the set W of worldview transformations between inertial observers satisfies W = G. This holds in particular for all relevant subgroups of Gal cPoi, and cEucl (the groups of Galilean, Poincaré and Euclidean transformations, respectively, where c ∈ Q is a model-specific parameter corresponding to the speed of light in the case of Poincaré transformations).In doing so, by an elementary geometrical proof, we demonstrate our main contribution: spatial isotropy is enough to entail that the set W of worldview transformations satisfies either W ⊆ Gal, W ⊆ cPoi, or W ⊆ cEucl for some c > 0. So assuming spatial isotropy is enough to prove that there are only 3 possible cases: either the world is classical (the worldview transformations between inertial observers are Galilean transformations); the world is relativistic (the worldview transformations are Poincaré transformations); or the world is Euclidean (which gives a nonstandard kinematical interpretation to Euclidean geometry). This result considerably extends previous results in this field, which assume a priori the (strictly stronger) special principle of relativity, while also restricting the choice of Q to the field R of reals.As part of this work, we also prove the rather surprising result that, for any G containing translations and rotations fixing the time-axis t, the requirement that G be a subgroup of one of the groups Gal, cPoi or cEucl is logically equivalent to the somewhat simpler requirement that, for all g ∈ G: g[t] is a line, and if g[t] = t then g is a trivial transformation (i.e. g is a linear transformation that preserves Euclidean length and fixes the time-axis setwise).2010 Mathematics Subject Classification. Primary 51P05 (83A05, 70B99); 20A15; 46B20.7 This can be expressed in our formal language without any assumption about the structure of quantities as: R(4 , here R( p ) 2 , R( p ) 3 , and R( p ) 4 denotes the second, third and fourth component of R( p ) ∈ Q 4 , i.e. if R( p ) = (t, x, y, z), then R( p ) 2 = x, R( p ) 3 = y, and R( p ) 4 = z.