A Philosophical Exploration of Shinuy: What Constitutes Change?

Examining the Talmud’s commitment to earlier rabbinic sources by exploring a b. Bava Kamma sugya about what constitutes change in stolen objects.

Dr. Ariel Furstenberg


One of the basic principles in the ninth chapter of m. Bava Kamma, governing the laws of robbery, is that if an article has changed since being stolen, the thief must pay the cash value of the article as it was at the time of the theft. This rule raises the question of what constitutes change (shinuy), specifically when an object undergoing change ceases being identical to its former self.

Theseus’ paradox
A famous philosophical debate on the issue of identity, known as the Ship of Theseus, can bring some of the issues into focus:

A ship made up of wooden planks is replaced one by one with other identically shaped planks.

If eventually all the original planks are replaced – is the current ship identical to the original one?

What if we took all the original planks and constructed another ship in the same exact manner as the original one – is this also identical to it?1

In this piece I present some of the Bavli’s discussion of m. Bava Kamma 9. Along with explaining the various talmudic conceptions of change, I will also consider the different forms of commitment by which a commentarial work, like the Bavli, relates to its earlier sources, and how such commitment influences the development of amoraic discourse.

The Nature of the Talmudic Commitment
to Earlier Rabbinic rulings

A typical talmudic discussion begins with a commentary on a mishnah, generally relating to some aspect of Jewish law. From there, the discussion branches out and can expand into many possible topics, some only tangentially related to the original discussion.2 During the process of deliberation, talmudic interlocutors often refer to earlier texts from the tannaitic period or from the even earlier biblical period.

At almost every stage of the discussion the talmudic speaker must choose3 the type and level of commitment he holds towards earlier rabbinic rulings. He determines whether he is committed to just using the same words that appeared earlier, or to the concepts behind these words, or to earlier conceptions containing these concepts.4 This defines the framework in which the talmudic speaker conducts his argumentations at a particular stage.

Being committed to past expressions mainly means that they become part of the ongoing later discourse; but in what sense do they integrate into later discourse? How do they influence the discourse?

The Mishnaic Approach – Natural Language

The first mishnah5 of the ninth chapter of tractate Bava Kamma concludes that all thieves have to make repayment כשעת הגזלה – “in accordance with the time of the robbery.” This means that if the stolen article has not undergone change since the time of the robbery, the thief is expected to return the article itself. But if the article has changed, the thief has to pay the cash value of the article as it was at the time of the theft.

As is often the case, the Mishna does not provide any explicit criterion of change and identity. It just presents us with examples of the principle.


Objects that have changed

משנה בבא קמא ט:ב

גזל בהמה והזקינה, עבדים והזקינו, משלם כשעת הגזלה….

גזל מטבע ונסדק, פרות והרקיבו, יין והחמיץ, משלם כשעת הגזלה.

m. Bava Kamma 9:2

If one has stolen an animal and it aged, or slaves and they aged, he must repay their cash value as at the time stolen…

If he has stolen a coin and it cracked, fruits and they rotted, wine and it became sour, he repays their cash value as at the time stolen.

Objects that have not changed
מטבע ונפסל, תרומה ונטמאת, חמץ ועבר עליו הפסח…אומר לו הרי שלך לפניך.
[However, if one has stolen] a coin which afterward became invalid, teruma and it became unclean, or chametz (leaven) which was in the hand of the robber during Passover… [in these cases] he may say: “Yours is before you.”

Besides the principle that change must be physical in order to be considered change, the Mishnah assumes that things change and that we know more or less when that happens.

The Bavli’s Explication:
The First Criterion: Reversibility

The Bavli, however, sees a need for precise definitions and criteria regarding change and identity.6 This is a general tendency in the Bavli, along with abstractions and generalizations.7 The first criterion the amoraim offer is reversibility (b. Bava Kamma 93b): If the change is reversible and we could return the article to its original form and state, the thief may return the article itself – even without actually reversing it. But if it is not reversible, we consider this to be change; the article is not identical to its original self and the thief has to pay its cash value.

This is best illustrated by a contemporary example: If someone steals an unassembled table from IKEA and puts it together, it is reversible, since he could just take it apart. In contrast, if one cuts wood into pieces it is not reversible.

Commitment to the Past
In the process of constructing the criterion of reversibility, the amoraim of the sugya are constantly trying to reconcile it with earlier sources, thereby refining their understanding and scope of the abovementioned criterion.8

In the following example the Bavli compares the ruling of our mishnah to another mishnah dealing with the law of “first of the fleece” (reishit ha’gez) provisions given to the priests. There too, if the fleece is processed and goes through change, it is exempt from going to the priest.  However, the criteria are inconsistent:

בבלי בבא קמא צג ע”ב

‘…גזל עצים ושיפן, אבנים וסיתתן, צמר וליבנו, פשתן ונקהו משלם כשעת הגזלה’.

וליבון מי הוי שינוי?

ורמינהי: ‘לא הספיק ליתנו לו עד שצבעו – פטור; לבנו ולא צבעו – חייב!’


רבא אמר:…לא קשיא: הא דנפציה נפוצי, הא דסרקיה סרוקי.

רבי חייא בר אבין אמר: הא דחווריה חוורי, הא דכבריה כברויי.

b. Bava Kamma 93b

“…one who stole pieces of wood and planed them, stones and chiseled them, wool and bleached it or flax and cleansed it, he repays their cash value as at the time stolen.”

But should bleaching be considered a change?

A contradiction arises from the following [mishna in Hulin]: “If the owner did not manage to give the first of the fleece to the priest until it had already been dyed, he would be exempt, but if he bleached it without having dyed it, he would still be liable.”


Rava said: …there is no difficulty, as in one case [the process of bleaching was] by beating the wool, whereas in the other case the wool was corded with a comb.

R. Hiyya b. Abin said that in one case the wool was merely washed whereas in the other it was whitened with sulfur.

Rava and R. Hiyya b. Abin’s solutions are to differentiate between various levels of bleaching. This leads to a similar comparison between tanaitic sources regarding wool which has been dyed in which Rava clearly differentiates between reversible and non-reversible dye.

One of the results of this commitment to earlier rulings and cases is the need of the amoraim to interpret the mishnaic examples according to their own criteria of reversibility. For example, R. Ashi explains our mishnah as follows:

‘עצים ועשאן כלים’ – בוכאני, דהיינו שיפן; ‘צמר ועשאן בגדים’ – נמטי, דהיינו שינוי דלא הדר(בבלי, בבא קמא צג ע”ב)

‘Wood made into utensils’ – meaning ‘into pestles’, which involve change done by carving the wood; ‘wool made into garments’ similarly means into felt clothes, which involves a change that does not revert to its original form. (b. Bava Kamma 93b)

The Second Criterion:
New Name

The second criterion that appears in the Talmud is that of a “new name”: Only if the stolen object is called a different name as a result of the physical change it is considered a change.

The incompatibility between this and the first criterion are evident in the following ruling by Rav Pappa:

בבלי, בבא קמא צו ע”א

אמר רב פפא: האי מאן דגזל דיקלא מחבריה וקטליה, אע”ג דשדיא מארעא לארעא דידיה – לא קני.

מאי טעמא? מעיקרא דיקלא מיקרי והשתא נמי דיקלא מיקרי.

דיקלא ועביד גובי– לא קני. השתא מיהת גובי דדיקלא מיקרי.

גובי ועבדינהו כשורי – קני.

כשורי רברבי ועבדינהו כשורי זוטרי – לא קני.

עבדינהו קצוצייתא – קני.

b. Bava Kamma 96a

R. Papa said: He who stole a palm-tree and cut it down, did not acquire title [=change has not occurred], although he removed it to his own field.

Why so? Because before it was called “palm-tree” and now it is called “palm-tree.”

[So also] where out of the palm-tree he made logs he would not acquire title to them, as even now they would still be called ‘logs of a palm-tree’.

It is only where out of the logs he made beams that he would acquire title to them.

But if out of big beams he made small beams he would not acquire title to them.

However, if he made them into boards he would acquire title to them.

This criterion is further explicated with the aid of a number of illustrations. An example discussed by Rava is that if one takes a lulav and turns it into a broom – that is a new name (from “lulav” to “broom”), hence considered a legally significant change.

A New Criterion Without Commitment to the Past
Though both first and second criteria provide us with a rather exact definition – especially in contrast to the Mishnah, which omits precise definitions entirely –  they do so in a different manner. Significantly, over the course of the development of the new name criterion, there is no reference to any past halakhic examples or concepts. All the cases provided to exemplify the criterion are artificial and, apparently, invented for the criterion’s own sake. The amoraim involved in the discussion show no commitment to the past tradition.

Moreover, the criterion contradicts the rulings / cases provided by the Mishnah as examples of change. For instance, the Mishnah discusses animals that have aged and rotten fruit.  Aging animals and rotten fruit are still “animals” and “fruit.”

The Third Criterion:
Panim Chadashot

The next stage of development in this sugya is the conception of panim chadashot, “new entities,” or literally “new face(s).” This conception provides a complex criterion, one that involves reversibility, on the one hand, and new name on the other. The following passage makes its case via two vivid illustrations.


Example 1

בבלי, בבא קמא צו ע”ב

אמר רב פפא: האי מאן דגזל עפרא מחבריה ועבדיה לבינתא – לא קני.

מאי טעמא? דהדר משוי ליה עפרא.

לבינתא ועבדיה עפרא – קני.

מאי אמרת? דלמא הדר ועביד ליה לבינתא!

האי לבינתא אחריתי הוא, ופנים חדשות באו לכאן.

b. Bava Kamma 96b

R. Papa said: if one stole sand from another and made a brick out of it – he would not acquire title to it [=change has not occurred].

The reason being that it could again be made into sand.

But if he converted a brick into sand he would acquire title to it.

Why is that so? Perhaps he could make the sand again into brick!

– [It may be said that] that brick is not the original but another brick and it is considered a “new face” (a new entity).

Example 2

ואמר רב פפא: האי מאן דגזל נסכא מחבריה ועביד זוזי– לא קני.

מאי טעמא? הדר עביד להו נסכא

זוזי ועבדינהו נסכא
– קני.

מאי אמרת? הדר עביד להו זוזי!

פנים חדשות באו לכאן.

R. Papa further said: if one stole a bar of silver from another and converted it into coins – he would not acquire title to them.

The reason being that it could again be made into a bar of silver.

But if out of [stolen] coins he made a bar of silver he would acquire title to it.

Why is that so? Perhaps he could make the bar into coins again!

– they will be considered a “new face” (a new entity).

The Talmud’s thinking is that by turning a brick into a pile of sand; the brick has lost its name “brick.” Hence, reversibility cannot help it retain its original status and it is considered a legally significant change. On the other hand, a “pile of sand” is not considered as having a name in the first place – its name is just a description of the sort of thing philosophers would call “stuff.”9 And when this type of stuff changes it does not lose its name, hence if it is reversible we consider it as not being changed at all.

A New Criterion out of Two Earlier Ones
The criterion of “new entity” is conceptually constructed with the aid of the two criteria developed in the earlier amoraic stages: reversibility and new name. However, the Talmudic sugya gives no reference to earlier halakhic rulings and cases, neither from this mishnah nor from any other source. The brick and coin examples given are provided to explain the criterion, but not taken from traditional sources and do not necessarily fit them.10 In this stage, we see commitment to concepts from the past, but not to rulings and halakhic cases from the past.


While studying the development of different conceptions of change and identity regarding stolen articles, we also see different commitments to the past at each stage.

  • The first criterion (“reversibility”) reflects a high level of commitment to past rulings and halakhic cases.
  • The second criterion (“new name”), attributed to another amoraic speaker, showed no commitment to anything in the past.
  • The final criterion (panim chadashot) was developed through commitment to past concepts but not to past halakhic rulings and cases.

These different commitments towards the past, presented by the amoraim at each stage, had an immediate effect on the resulting development.

This phenomenon is equivalent, in a way, to changing the rules of a game while playing or having each player play according to different commitments and rules. Such a development might be surprising, but it is not uncommon in Talmudic discourse.

These different commitments resulted in different constraints that are imposed on the conceptual development. In general, one’s implicit and explicit commitments create a framework to which he or she is bound; and one’s conceptual development is constrained by this framework. It is a deep and practical philosophical question to consider the way one is bound by such a framework and how it constrains him or her.

In our context, we can see that had the amoraic speaker proposing the criterion of new name been committed to earlier halakhic rulings and cases, he could not have developed the idea of new name as he did, since it does not fit the rulings of the Mishnah. The rulings and cases in the Mishnah would have hindered such a possibility; either this possibility would not have even come up within this amora’s framework, or it would have required using strong interpretational force.

Evolution and Innovation as a Factor
in the Sugya itself

Finally, it is worth noting that this sugya can be considered a “mirroring sugya,” in the sense that the sugya’s subject matter has a parallel, related “meta” issue which we are interested in:11 While discussing the issue of change and identity that arises in the Bavli in the context of theft, we also considered the issue of halakhic evolution itself. In encountering this meta issue, we were encouraged to consider whether we are witnessing change in halakha or change in the laws of robbery; and whether change in the laws of robbery is an example of progression or innovation.



Dr. Ariel Furstenberg completed his graduate studies at the Cohn Institute for the History and Philosophy of Science at Tel Aviv University, focusing on concept formation and conceptual change, specifically in relation to the Babylonian Talmud. He also held a research position at The School of Social Science at The Institute for Advanced Study, Princeton, and is currently a research associate at the Edmond and Lily Safra Center for Brain Sciences (ELSC) at the Hebrew University of Jerusalem. His current interests lie on the interface between philosophy and brain science, specifically regarding issues of freedom, will and agency. He is the author of the forthcoming (2016) book: The Languages of Talmudic Discourse: A Philosophical Study of the Evolution of Amoraic Halakha (Magnes Press and Van Leer Jerusalem Institute Press). 

  1. See discussion in Eli Hirsch, The Concept of Identity (Oxford, 1992), pp. 68-69.
  2. For a discussion of tangents in the Bavli and their possible connection to orality, see Shai Secunda’s TABS essay, “Why the Talmud is the only Rabbinic Work from Babylonia.”
  3.  I use the term “choose” although it is not always clear whether it is a conscious or unconscious choice.
  4. I follow here the distinction emphasized by Hilary Putnam between concepts and conception. See Hilary Putnam, Reason, Truth and History (Cambridge, 1981), pp. 103-126. In brief, a conception is an underlying framework determining the manner in which the concepts relate to each other and to the world. It is parallel to what we would term a ‘world-view’ or ‘paradigm’ as opposed to just ‘concepts.’ 
  5.  A word on my methodology: It is a well-known fact that the lion’s share (or maybe all) of the deliberations that appear in the Talmud constitute retrospective reconstructions. Nevertheless, even if the amoraic reconstruction is not loyal to a historical truth, it certainly represents a possible process of halakhic development, in any case one that is learned and accepted as part of the Jewish formative canon (as opposed to being just a normative canon, to use Moshe Halbertal’s terminology. See Moshe Halbertal, People of the Book: Canon, Meaning, and Authority {Cambridge, 1997}). In this piece I am not concerned with the historical accuracy of this development; rather, I describe it as perceived in the eyes of those actively conducting the talmudic-halakhic discourse as a possible process of halakhic development.
  6. This process has been termed “explication” by philosopher Rudolf Carnap. See his Logical Foundations of Probability, (Chicago, 1950).
  7.  See L. Moscovitz, Talmudic Reasoning: From Casuistics to Conceptualization (Tubingen, 2002); and  J. L Rubenstein, “On Some Abstract Concepts in Rabbinic Literature”, Jewish Studies Quarterly 4 (1997), pp. 33–73
  8. In addition, defining change as linked to reversibility gives the amoraim a powerful tool for dealing with many more traditional sources, even those not related to theft.
  9. One might consider this as the difference between naming objects and descriptions which are different methods of sorting the world around us. See the article on “Sortals” in the Stanford Encyclopedia of Philosophy (http://plato.stanford.edu/entries/sortals/). This distinction is also articulated via the mass/count distinction: between nouns that we ask “haw many?” and those that we ask “how much?”. Compare to E. Hirsch, “Identity in the Talmud”, Midwest Studies in Philosophy 23 (1999), pp. 166—180.
  10. In many cases the Tosafot try to bridge this gap by interpreting the new conception in light of the earlier mishnaic sources; see for example tosafot on b. Bava Kamma 96b “ועבדיה זוזי לא קני”.
  11. I thank Avi Warshavsky for the suggesting the term “mirroring” in this context. 
Print Friendly