Handwritten text is both shape and meaning. In the writing process, the scribe has to combine different kinds of knowledge and aims, their graphical training and competence, grammatical understanding, comprehension of the meaning, and the need to pass it on to the readers. Word spacing is part of the scribe’s tools. The present contribution reassesses the definition and measures of space, to promote an objective measure while space is at the same time “seen” and “read”. It confirms the importance of Paul Saenger’s “unity of space” based on the space between minims rather than on the height of minims, and it introduces a comparison with the average character width and a measurement of the “blank area” between words. Indeed, as part of the Oriflamms research project, a tool has been developed in order to automate the measurements of digital images. This preliminary research examines forty-eight manuscripts and demonstrates that spacing is not a binary practice, with agglutinated words on the one hand and regularly separated words on the other. Grammatical categories influence how words are separated and, if we can confirm Paul Saenger’s observation on “fere canonical separation” with agglutinated prepositions, there is most notably a full range of variation. According to the meaning of words, the space between them can be shorter or wider. In the measurements and their interpretations, one must be extremely cautious with regard to allographs and shapes, abbreviations and semiotic ambiguity, because text is shape and meaning at the same time. Further research is needed. We suggest using the average character width to normalize observations on space length. Additional measurements can be explored to transform measures in pixels based on digital images into more meaningful ratios compared with manuscript size, the interlinear space, and the script proportions, and to provide a normalizing method for the blank area. To compare pairs of manuscripts, a deeper analysis into scribal profiles based on the analysis of each manuscript’s specific figures could overcome the diverse dimensionalities. The present observations on the conjunction “et” are a clear indication that the study of further allographs and abbreviations would be needed. From a methodological perspective, we can now extend the inquiry towards inner-word spacing and compare this with medieval grammatical theories and phonology, especially regarding the space that appears within a verb after a prepositional prefix. The complex interdependence between grammar, paleography, and the continuum in the range of space sizes reveals a paradoxical challenge: how to threshold and re-binarize the results to decide whether or not a scribe intended to communicate a space between two different words.