Have an Access Token?
Enter your access token to activate and access content online.
Please login and go to your personal user account to enter your access token.
Have Institutional Access?
Access content through your institution. Any other coaching guidance?
Don't have an account?
Figure 1. This figure shows the system of German definite articles in three different time periods: Old High German (OHG; 900–1100; Wright, 1906), Middle High German (MHG; 1100–1500; Wright, 1916), and New High German (NHG; from 1500 onwards). Over the course of time, multiple forms have collapsed into syncretic articles. Grey cells indicate forms that have changed with respect to the previous time period. In the OHG and MHG systems, I followed Wright’s convention to use ë for indicating the short vowel [ɛ] and ē for the long vowel [eː] (which is the pronunciation of the letter e in all NHG definite articles except die).
Figure 2. This chart compares Old High German (black) and New High German (white) with regard to how many utterances were disambiguated during parsing in four different analyses. In all four sets of results, the agents had access to the information carried by the determiners and the number-gender information from their head nouns (
). The first set of results shows that, in isolation, the OHG determiners are more reliable for disambiguating utterances than the NHG determiners. However, as seen in the other sets of results, where the listener is also allowed to exploit other grammatical cues such as subject-verb agreement (
) and/or selection restrictions (
), the difference between both systems almost disappears.
Figure 3. These spider charts each take an article at their center and then show the interpolated distances between that article and other forms of the same paradigm. The center articles are OHG die (top left) and dëru (bottom left) and their corresponding forms in NHG die (top right) and der (bottom right). When comparing the spider charts for OHG and NHG die, we see that a cluster of close OHG forms (die, diu and deo) have collapsed into a single form in NHG, which has improved the auditory distinctiveness of the paradigm. When comparing OHG dëru to NHG der, we see that the OHG forms of dëmu and dëru have undergone phonological reduction, which results in three close neighbors in the NHG system: der, den and dem. Despite their low auditory distinctiveness, these forms have been maintained in order to uphold the language’s disambiguation power.
Figure 4. This figure summarizes the performance profiles of the OHG system (black) and the NHG system (white). The OHG system offers more cue reliability, but the overall disambiguation power of both OHG and NHG is almost equivalent. In all other linguistic assessment criteria (processing effort, ease of articulation and auditory distinctiveness), the NHG paradigm outperforms its OHG predecessor.
Character limit 500/500