The aim of this article is to demonstrate the useful role that can be played by spectropolarimetric observations of young and evolved emission line stars that analyse the linearly polarized component in their spectra. At the time of writing, this demonstration has to be made on the basis of optical data since there is no common-user infrared facility, in operation, that offers the desired combination of spectral resolution and sensitivity. Here we focus on what can be learned from linear spectropolarimetry alone at reasonably high spectral resolution and at $10^3 < $S/N$ < 10^4$. And we remind that the near infrared (1--2 micron) has the potential to out-perform the optical as a domain to work in because of the greatly reduced interstellar obscuration at these wavelengths. This point has been reached at a time when theory, exploiting flexible Monte Carlo methods, is fast becoming a powerful tool. In short we have the complex phenomena, and the rise of the modelling capability to match -- good data are the missing link.

11 pages, ESO Conference on High Resolution Infrared Spectroscopy