In churches which employ a pastor's assistant, deaconess, or church visitor, this worker often knows more of the things that the case worker seeks information about than the minister does.

2. To Supply the Results of Their Inquiries. There is a second function of social agencies as witnesses; namely, to supply those objective data of a fundamental sort about clients which seldom change. Gathered originally by a particular organization for its own purposes, these facts may, if collected carefully in the first place, serve equally well, and with substantial economy of time and effort, the purposes of a second organization. A more important saving, even, is the wear and tear to the sensibilities of a client who might otherwise be harassed by needless questioning.1

It will help us to evaluate the evidence of social agencies more justly if we can keep quite distinct in our own minds these two functions of theirs as witnesses. In their testimony from personal experience they are witnesses of their own knowledge; in testimony based upon data that they have gathered they are often witnesses to matters of hearsay. Even in this second capacity, they have the advantage sometimes, though not always, of a certain skill in the weighing and testing of evidence. If their items of outside evidence have been recorded and duly credited to their sources at the time that they were gathered with sufficient fullness and accuracy to make them clear, and if the written record is always consulted before reporting to another agency, the danger of error is materially reduced. Danger of error in the original observation remains, however, and no one agency is ever an equally good judge of all kinds of data. Agencies that habitually neglect certain observations may be very keen about certain others. Such differences can be discovered by practical experience only. The particular type of social work engaged in, with its natural limitations, is one guide; the history not only of the individual agency but of the form of effort of which it is a part is another; the past relations of the given agency to the one seeking information is still another; but the most important factor of all is the native ability and professional equipment of the individual case worker who made the original observation or who represents the agency as witness, and this is a changing factor.

1 See also the discussion of duplicate investigations on p. 311.

A case worker who removes from one city to another must revise all his standards of measure of social agency testimony. A charity organization society or associated charities, for example, is usually an agency that thinks of and knows a family as a whole. Relatively speaking, it has an unusually clear idea of the family histories and background of its clients, is well grounded in the habit of conferring with relatives, health agencies, former employers, schools, and the social agencies interested. It is not so strong in ability to gauge neighborhood influences; it too often neglects to individualize each growing child in the family; it sometimes emphasizes health and self-support at the expense of less measurable but very important social gains. But in certain cities the societies bearing the name of charity organization society or of associated charities are only rudderless, small-dole agencies, operating without plan or purpose. Obviously, it is necessary to look beyond the name and the avowed objects of a society in accepting its testimony.

A case worker may have been well trained, but may be employed by an agency under conditions that make it impossible for him to do trained work. These conditions limit his competence as a witness, of course. Some public departments, for instance, investigate chiefly with reference to the question of ability or inability to reimburse the state, or with reference to legal settlement. Others have rules that so many references-three or four-are to be consulted. Limitations of this order must be known and allowed for.

A radical change of management in an agency often makes it necessary to note whether a particular investigation, together with the inferences and plan of treatment drawn from it, was made before the change or after. One of the case readers for this volume spent two months in a society to protect children from cruelty in which there had been a change of management. After reading a large number of its records, she wrote:

"Even those cases in which only one or two or three interviews are thought to be necessary show an absolutely new way of approach since the change. I think I mentioned before that, under the old régime, unless the evidence of the investigator's eyes, backed up, perhaps, by a policeman or a neighbor, showed obvious neglect, the entry 'Nothing for us to do' was by no means uncommon. I have not found one case, under the new régime, where this sentiment is expressed in letter or in spirit. There is always something for the S. P. C. C. to do, though this is not always taking the case into court."

This last illustration might also point the moral that agencies skilled in the gathering and recording of objective data are usually the ones most likely to continue treatment long enough to gather also a rich store of personal experiences. This is not always true, however. Resourceless treatment, together with inability to recognize the significance of reactions to treatment, may follow a fairly good social diagnosis. It should be repeated, therefore, that the two types of evidence—as to actual, first-hand experiences with clients, and as to objective data gathered outside about them

-must be distinguished, and that from each social agency must be had the type of witness that it is best able to give. Social agencies are not what Francis H. McLean calls a "single-headed" source. They are so far from being this that they presented, in the material studied for this book, examples of the very best and of the very worst social reporting.

With the increasing activity of social organizations and with their marked tendency toward specialization, has come a new need of some systematic exchange of information. This has become a prime necessity, if only to fix responsibility for social treatment, though it has many other advantages. As we have seen in the chapter on The First Interview, the first step, upon receipt of an application from a new client, is to discover what other social agencies, if any, are acquainted with him. This discovery is facilitated by the regular exchange of information among agencies, or, better still, by an exchange of the identifying data which will lead us to sources of information. Very easily and conveniently these identifying memoranda lead to an exchange of information when it is needed and not otherwise.

II. THE CONFIDENTIAL EXCHANGE

Some years ago the writer had occasion to consult an oculist in a strange city. He proved very skilful, so that when she removed her home from Philadelphia to New York and needed to choose another regular oculist, she asked the advice of this one. Upon her first visit to the New York practitioner thus selected, she tried to give the history of former eye treatments, first in her old home, then in the city only visited, and so on, making a sincere effort to do this as accurately as she could. But the new doctor received these communications with an air of skepticism. It appeared that he had been supplied with more trustworthy data already; the specialist who had recommended him had sent him a detailed statement of the Philadelphia prescriptions, and of his own findings and prescriptions. These were more truthful than the oral witness of the patient for two reasons: they were taken from memoranda made at the time, and, more important, they were taken from memoranda made by expert refractionists.

1 See p. 175.

Such communication direct from practitioner to practitioner greatly reduces the percentage of error in every profession, and upon the need of some such direct interchange of experiences in social work is based a system now widely adopted, especially in our larger cities, and already referred to in these pages more than once. It was known first as the "registration bureau" and later as the "confidential exchange," or the "social service exchange." Its use brings other minor advantages, such as a checking of the tendency (not confined to the clients of social agencies) to run around from adviser to adviser. Doctors, lawyers, architects, and many other professional men, probably, know the type of nervous indecision, of speculative fever, of which such running around is only a symptom.

The Confidential Exchange was devised by the charity organization societies and is still financially supported or administered or both by these societies in most places. Better diagnosis, better treatment, better understandings among agencies, are its outstanding achievements, but incidentally it has reduced duplication of effort, has increased the sense of responsibility of the social agency definitely in charge of the individual case, and has been, moreover, a real economy. Its advantages are not confined to the processes of social diagnosis, for, long after treatment has begun, an inquiry at the Exchange from another agency about one of our clients may enable us to prevent unnecessary interference or may assure much needed co-operation; but the present account of the Exchange is limited to its uses in diagnosis.

Wherever there has been no Exchange and then one is established, its possible usefulness is soon vividly illustrated by such instances as the following, taken from Miss Margaret F. Byington's pamphlet study of Exchanges:1

In another city a Confidential Exchange is just being started, and the infant mortality nurses and the tuberculosis nurses have not yet learned to use it. One family was badly infected with tuberculosis, the father dying, and the mother in an advanced stage of the disease. There were seven children, the youngest a nursing baby. The tuberculosis nurse kept urging the mother to stop nursing the child, but she refused to do so. Finally the tuberculosis nurse found that the infant mortality nurse had been visiting the family and, not knowing that the mother had 1 The Confidential Exchange, p. 8.

tuberculosis, was insisting that she nurse the child. When the two nurses got together on the case, it was too late, for the baby died of tubercular meningitis.

Here was failure to consult owing to ignorance of the other agency's relation, but our case reading shows many instances of such failure where the other agency's relation was known. This sense of self-sufficiency, this tendency to operate in a vacuum, is worn down by the successfully administered Exchange. Not all who have the advantage of a local Exchange use it, as the case just cited also shows, but a few such happenings as this one (and oversights quite as serious are brought to light as soon as an Exchange is started) convey their own lesson.

The following brief description of Confidential Exchange machinery is Miss Byington's:1

"The mechanism of the Exchange is an alphabetical index with a card for each family or unattached person known to any of the inquiring agencies. This card gives the 'identifying information' -the names, ages, and occupations of the members of the family group, names and addresses of relatives, and the names of agencies interested, with the date on which each inquired. No facts about family history or treatment are included. When a co-operating society becomes interested in a new family, or in any one of its members, it inquires at once whether the Confidential Exchange knows the family or person. This inquiry is made either by telephone or by mail on printed slips furnished by the Exchange. The Exchange looks up the family in the index, and then reports to the inquiring agency the names of any societies that have been previously interested and the dates on which they inquired. If the information given by the inquirer is not sufficient to make identification possible, the agency is so notified, with the request that it inform the Exchange when further facts are secured.2 The Chil1 The Confidential Exchange, p. 5 sq.

The following on the subject of identifying information is part of a longer passage in Wigmore's Principles of Judicial Proof, pp. 64-65. "The process of constructing an inference of identification consists usually in adding together

a number of circumstances, each of which by itself might be a feature of many objects, but all of which together can conceivably coexist in a single object only. Each additional circumstance reduces the chances of there being more than one object so associated. It may be illustrated by the ordinary case of identification by name. Suppose there existed a parent named John Smith, whose heirs are sought; and there is also a claimant whose parent's name was John Smith. The name John Smith is associated with so many persons that the chances of two