now lets take a pair of words like        « gene­rous » and « thrif­ty » say we could pro­ba­bly find an axis        that ran through them unfor­tu­na­te­ly we could find many more        than one axis but lets take an axis an axis is a good        word it sug­gests so much a kind of space through with        it runs a kind of seman­tic globe domain ? hyper­space ?        any­way lets call it an axis        […]        but what could that mean that        a word could lie clo­ser to the same axis than ano­ther it could        mean that we will have to find only pure oppo­sites or anto­nyms        lying at ends of fea­ture axes and that all the words in the        sys­tem will have to be plot­ted by the inter­sec­tion of various axes        their spa­tial coor­di­nates in some kind of hyper­space so that        we know just how far off the axis of « clo­sed » and « open »        « gene­rous » and « thrif­ty » may real­ly be and how close they lie        to an axis of « big » and « small » for example or « soft » and        « hard »        how many fea­tures will we need to map any lexeme        how many fea­tures are there is there a fea­ture axis that        can be construc­ted by dra­wing a line bet­ween any two words        that can be regar­ded as oppo­sites seen from some point of        view will we have to connect eve­ry word with eve­ry        other word in prac­tice in prin­ciple dol­lars and dough­nuts        if not will        there be a finite set of such contrasts ?        […]        now the rea­son i chose to talk about tuning        i was pro­po­sing a way of loo­king at how we unders­tand        things how we come to unders­tand things come to an        unders­tan­ding with each other about things through        lan­guage has some­thing to do with a notion pro­cess        i would like to call tuning