The wave functions of the nuclei 14C and 14N were calculated using shell model matrix elements obtained from the Hamada-Johnston potential. This interaction has a tensor part and fits two nucleon scattering data to several hundred MeV. Provided one is allowed to vary the p 1 2-p 3 2 single particlesplitting,one can explain the fact that the β transition rate from 14C to 14N is very nearly zero. A p 1 2-p 3 2 splitting of 9.2 MeV is required as compared with the experimental value of 6.3 MeV. With the central Kallio-Kolltveit interaction, it is not possible to get a vanishing transition rate, no matter what value of the spin orbit splitting is chosen. The Hamada-Johnston interaction leads to a very reasonable result for the very fast transition 18Ne → 18F - log ft = 2.92 as compared with the experimental value 3.05.