From John Dewey’s Logic: The The­ory of Inquiry, by way of John J. McDermott’s The Phi­los­o­phy of John Dewey: The Struc­ture of Expe­ri­ence, this sum­mary of Dewey’s own chap­ter on the nature of inquiry.

In par­tic­u­lar, this strikes me as some­thing that bears on many dis­cus­sions I’ve had about machine learn­ing and mod­ern sta­tis­tics. And it reminds me of a cul­tural prob­lem I’ve been wrestling with among genetic pro­gram­ming researchers and oper­a­tions research peo­ple for some time. And would be use­ful in explain­ing the ped­a­gogy and prac­tice of engi­neer­ing “crafts­man­ship”, and more specif­i­cally that of soft­ware development.

Oh, and com­plex sys­tems research and emer­gence, too. That’s in there, somehow.

So you can see why I might think it’s impor­tant to understand.

I can’t quite put my fin­ger on it, but some­thing in here—perhaps obfus­cated by what today we might per­ceive as a dif­fi­cult style, but which is an attempt to con­vey very spe­cific con­cepts in a way that tries to avoid misunderstanding—is vital to many threads in mod­ern life. In par­tic­u­lar, some­thing deeply impor­tant hap­pens down in the last para­graph, where I’ve high­lighted it.

I would love to have a cor­re­spon­dent who could dis­cuss this pro­duc­tively. Per­haps one might be found to read the orig­i­nal Dewey, or even the few sur­round­ing pages extracted in McDermott’s sum­mary, and tell me just what it is I’m respond­ing to?

…Inquiry is the directed or con­trolled trans­for­ma­tion of an inde­ter­mi­nate sit­u­a­tion into a deter­mi­nately uni­fied one. The tran­si­tion is achieved by means of oper­a­tions of two kinds which are in func­tional cor­re­spon­dence with each other. One kind of oper­a­tions deals with ideational or con­cep­tual subject-​​matter. This subject-​​matter stands for pos­si­ble ways and ends of res­o­lu­tion. It antic­i­pates a solu­tion, and is marked off from fancy because, or, in so far as, it becomes oper­a­tive in insti­ga­tion and direc­tion of new obser­va­tions yield­ing new fac­tual mate­r­ial. The other kind of oper­a­tions is made up of activ­i­ties involv­ing the tech­niques and organs of obser­va­tion. Since these oper­a­tions are exis­ten­tial they mod­ify the prior exis­ten­tial sit­u­a­tion, bring into high relief con­di­tions pre­vi­ously obscure, and rel­e­gate to the back­ground other aspects that were at the out­set con­spic­u­ous. The ground and cri­te­rion of the exe­cu­tion of this work of empha­sis, selec­tion and arrange­ment is to delimit the prob­lem in such a way that exis­ten­tial mate­r­ial may be pro­vided with which to test the ideas that rep­re­sent pos­si­ble modes of solu­tion. Sym­bols, defin­ing terms and propo­si­tions, are nec­es­sar­ily required in order to retain and carry for­ward both ideational and exis­ten­tial subject-​​matters in order that they may serve their proper func­tions in the con­trol of inquiry. Oth­er­wise the prob­lem is taken to be closed and inquiry ceases.

One fun­da­men­tally impor­tant phase of the trans­for­ma­tion of the sit­u­a­tion which con­sti­tutes inquiry is cen­tral in the treat­ment of judge­ment and its func­tions. The trans­for­ma­tion is exis­ten­tial and hence tem­po­ral. The pre-​​cognitive unset­tled sit­u­a­tion can be set­tled only by mod­i­fi­ca­tion of its con­stituents. Exper­i­men­tal oper­a­tions change exist­ing con­di­tions. Rea­son­ing, as such, can pro­vide means for effect­ing the change of con­di­tions but by itself can­not effect it. Only exe­cu­tion of exis­ten­tial oper­a­tions directed by an idea in which rati­o­ci­na­tion ter­mi­nates can bring about the re-​​ordering of envi­ron­ing con­di­tions required to pro­duce a set­tled and uni­fied sit­u­a­tion. Since this prin­ci­ple also applies to the mean­ings that are elab­o­rated in sci­ence, the exper­i­men­tal pro­duc­tion and re-​​arrangement of phys­i­cal con­di­tions involved in nat­ural sci­ence is fur­ther evi­dence of the unity of the pat­tern of inquiry. The tem­po­ral qual­ity of inquiry means, then, some­thing quite other than that the process of inquiry takes time. It means that the objec­tive subject-​​matter of inquiry under­goes tem­po­ral modification.

Ter­mi­no­log­i­cal. Were it not that knowl­edge is related to inquiry as a prod­uct to the oper­a­tions by which it is pro­duced, no dis­tinc­tions requir­ing spe­cial dif­fer­en­ti­at­ing des­ig­na­tions would exist. Mate­r­ial would merely be a mat­ter of knowl­edge or of igno­rance and error; that would be all that could be said. The con­tent of any given propo­si­tion would have the val­ues “true” and “false” as final and exclu­sive attrib­utes. But if knowl­edge is related to inquiry as its war­rantably assert­ible prod­uct, and if inquiry is pro­gres­sive and tem­po­ral, then the mate­r­ial inquired into reveals dis­tinc­tive prop­er­ties which need to be des­ig­nated by dis­tinc­tive names. As under­go­ing inquiry, the mate­r­ial has a dif­fer­ent log­i­cal import from that which it has as the out­come of inquiry. In its first capac­ity and sta­tus, it will be called by the gen­eral name subject-​​matter. When it is nec­es­sary to refer to subject-​​matter in the con­text of either obser­va­tion or ideation, the name con­tent will be used, and, par­tic­u­larly on account of its rep­re­sen­ta­tive char­ac­ter, con­tent of propositions.

The name objects will be reserved for subject-​​matter so far as it has been pro­duced and ordered in set­tled form by means of inquiry; pro­lep­ti­cally, objects are the objec­tives of inquiry. The appar­ent ambi­gu­ity of using “objects” for this pur­pose (since the word is reg­u­larly applied to things that are observed or thought of) is only appar­ent. For things exist as objects for us only as they have been pre­vi­ously deter­mined as out­comes of inquiries. When used in car­ry­ing on new inquiries in new prob­lem­atic sit­u­a­tions, they are known as objects in virtue of prior inquiries which war­rant their assert­ibil­ity. In the new sit­u­a­tion, they are means of attain­ing knowl­edge of some­thing else. In the strict sense, they are part of the con­tents of inquiry as the word con­tent was defined above. But ret­ro­spec­tively (that is, as prod­ucts of prior deter­mi­na­tion in inquiry) they are objects.

[Lat­ter empha­sis is mine.]