Elongated magnetic polarities are observed during the emergence phase of bipolar active regions (ARs). These extended features, called magnetic tongues, are interpreted as a consequence of the azimuthal component of the magnetic flux in the toroidal flux-tubes that form ARs. We develop a new systematic and user-independent method to identify AR tongues. Our method is based on determining and analyzing the evolution of the AR main polarity inversion line (PIL). The effect of the tongues is quantified by measuring the acute angle [ tau] between the orientation of the PIL and the direction orthogonal to the AR main bipolar axis. We apply a simple model to simulate the emergence of a bipolar AR. This model lets us interpret the effect of magnetic tongues on parameters that characterize ARs ( e.g. the PIL inclination and the tilt angles, and their evolution). In this idealized kinematic emergence model, tau is a monotonically increasing function of the twist and has the same sign as the magnetic helicity. We systematically apply our procedure to a set of bipolar ARs that were observed emerging in line-of-sight magnetograms over eight years. For most of the cases studied, the tongues only have a small influence on the AR tilt angle since tongues have a much lower magnetic flux than the more concentrated main polarities. From the observed evolution of tau, corrected for the temporal evolution of the tilt angle and its final value when the AR is fully emerged, we estimate the average number of turns in the subphotospherically emerging flux-rope. These values for the 41 observed ARs are below unity, except for one. This indicates that subphotospheric flux-ropes typically have a low amount of twist, i.e. highly twisted flux-tubes are rare. Our results demonstrate that the evolution of the PIL is a robust indicator of the presence of tongues and constrains the amount of twist in emerging flux-tubes